""" Augmenters that somehow change the size of the images. List of augmenters: * :class:`Resize` * :class:`CropAndPad` * :class:`Crop` * :class:`Pad` * :class:`PadToFixedSize` * :class:`CenterPadToFixedSize` * :class:`CropToFixedSize` * :class:`CenterCropToFixedSize` * :class:`CropToMultiplesOf` * :class:`CenterCropToMultiplesOf` * :class:`PadToMultiplesOf` * :class:`CenterPadToMultiplesOf` * :class:`CropToPowersOf` * :class:`CenterCropToPowersOf` * :class:`PadToPowersOf` * :class:`CenterPadToPowersOf` * :class:`CropToAspectRatio` * :class:`CenterCropToAspectRatio` * :class:`PadToAspectRatio` * :class:`CenterPadToAspectRatio` * :class:`CropToSquare` * :class:`CenterCropToSquare` * :class:`PadToSquare` * :class:`CenterPadToSquare` * :class:`KeepSizeByResize` """ from __future__ import print_function, division, absolute_import import re import functools import numpy as np import cv2 import imgaug as ia from imgaug.imgaug import _normalize_cv2_input_arr_ from . import meta from .. import parameters as iap def _crop_trbl_to_xyxy(shape, top, right, bottom, left, prevent_zero_size=True): if prevent_zero_size: top, right, bottom, left = _crop_prevent_zero_size( shape[0], shape[1], top, right, bottom, left) height, width = shape[0:2] x1 = left x2 = width - right y1 = top y2 = height - bottom # these steps prevent negative sizes # if x2==x1 or y2==y1 then the output arr has size 0 for the respective axis # note that if height/width of arr is zero, then y2==y1 or x2==x1, which # is still valid, even if height/width is zero and results in a zero-sized # axis x2 = max(x2, x1) y2 = max(y2, y1) return x1, y1, x2, y2 def _crop_arr_(arr, top, right, bottom, left, prevent_zero_size=True): x1, y1, x2, y2 = _crop_trbl_to_xyxy(arr.shape, top, right, bottom, left, prevent_zero_size=prevent_zero_size) return arr[y1:y2, x1:x2, ...] def _crop_and_pad_arr(arr, croppings, paddings, pad_mode="constant", pad_cval=0, keep_size=False): height, width = arr.shape[0:2] image_cr = _crop_arr_(arr, *croppings) image_cr_pa = pad( image_cr, top=paddings[0], right=paddings[1], bottom=paddings[2], left=paddings[3], mode=pad_mode, cval=pad_cval) if keep_size: image_cr_pa = ia.imresize_single_image(image_cr_pa, (height, width)) return image_cr_pa def _crop_and_pad_heatmap_(heatmap, croppings_img, paddings_img, pad_mode="constant", pad_cval=0.0, keep_size=False): return _crop_and_pad_hms_or_segmaps_(heatmap, croppings_img, paddings_img, pad_mode, pad_cval, keep_size) def _crop_and_pad_segmap_(segmap, croppings_img, paddings_img, pad_mode="constant", pad_cval=0, keep_size=False): return _crop_and_pad_hms_or_segmaps_(segmap, croppings_img, paddings_img, pad_mode, pad_cval, keep_size) def _crop_and_pad_hms_or_segmaps_(augmentable, croppings_img, paddings_img, pad_mode="constant", pad_cval=None, keep_size=False): if isinstance(augmentable, ia.HeatmapsOnImage): arr_attr_name = "arr_0to1" pad_cval = pad_cval if pad_cval is not None else 0.0 else: assert isinstance(augmentable, ia.SegmentationMapsOnImage), ( "Expected HeatmapsOnImage or SegmentationMapsOnImage, got %s." % ( type(augmentable))) arr_attr_name = "arr" pad_cval = pad_cval if pad_cval is not None else 0 arr = getattr(augmentable, arr_attr_name) arr_shape_orig = arr.shape augm_shape = augmentable.shape croppings_proj = _project_size_changes(croppings_img, augm_shape, arr.shape) paddings_proj = _project_size_changes(paddings_img, augm_shape, arr.shape) croppings_proj = _crop_prevent_zero_size(arr.shape[0], arr.shape[1], *croppings_proj) arr_cr = _crop_arr_(arr, croppings_proj[0], croppings_proj[1], croppings_proj[2], croppings_proj[3]) arr_cr_pa = pad( arr_cr, top=paddings_proj[0], right=paddings_proj[1], bottom=paddings_proj[2], left=paddings_proj[3], mode=pad_mode, cval=pad_cval) setattr(augmentable, arr_attr_name, arr_cr_pa) if keep_size: augmentable = augmentable.resize(arr_shape_orig[0:2]) else: augmentable.shape = _compute_shape_after_crop_and_pad( augmentable.shape, croppings_img, paddings_img) return augmentable def _crop_and_pad_kpsoi_(kpsoi, croppings_img, paddings_img, keep_size): # using the trbl function instead of croppings_img has the advantage # of incorporating prevent_zero_size, dealing with zero-sized input image # axis and dealing the negative crop amounts x1, y1, _x2, _y2 = _crop_trbl_to_xyxy(kpsoi.shape, *croppings_img) crop_left = x1 crop_top = y1 shape_orig = kpsoi.shape shifted = kpsoi.shift_( x=-crop_left+paddings_img[3], y=-crop_top+paddings_img[0]) shifted.shape = _compute_shape_after_crop_and_pad( shape_orig, croppings_img, paddings_img) if keep_size: shifted = shifted.on_(shape_orig) return shifted def _compute_shape_after_crop_and_pad(old_shape, croppings, paddings): x1, y1, x2, y2 = _crop_trbl_to_xyxy(old_shape, *croppings) new_shape = list(old_shape) new_shape[0] = y2 - y1 + paddings[0] + paddings[2] new_shape[1] = x2 - x1 + paddings[1] + paddings[3] return tuple(new_shape) def _crop_prevent_zero_size(height, width, crop_top, crop_right, crop_bottom, crop_left): remaining_height = height - (crop_top + crop_bottom) remaining_width = width - (crop_left + crop_right) if remaining_height < 1: regain = abs(remaining_height) + 1 regain_top = regain // 2 regain_bottom = regain // 2 if regain_top + regain_bottom < regain: regain_top += 1 if regain_top > crop_top: diff = regain_top - crop_top regain_top = crop_top regain_bottom += diff elif regain_bottom > crop_bottom: diff = regain_bottom - crop_bottom regain_bottom = crop_bottom regain_top += diff crop_top = crop_top - regain_top crop_bottom = crop_bottom - regain_bottom if remaining_width < 1: regain = abs(remaining_width) + 1 regain_right = regain // 2 regain_left = regain // 2 if regain_right + regain_left < regain: regain_right += 1 if regain_right > crop_right: diff = regain_right - crop_right regain_right = crop_right regain_left += diff elif regain_left > crop_left: diff = regain_left - crop_left regain_left = crop_left regain_right += diff crop_right = crop_right - regain_right crop_left = crop_left - regain_left return ( max(crop_top, 0), max(crop_right, 0), max(crop_bottom, 0), max(crop_left, 0)) def _project_size_changes(trbl, from_shape, to_shape): if from_shape[0:2] == to_shape[0:2]: return trbl height_to = to_shape[0] width_to = to_shape[1] height_from = from_shape[0] width_from = from_shape[1] top = trbl[0] right = trbl[1] bottom = trbl[2] left = trbl[3] # Adding/subtracting 1e-4 here helps for the case where a heatmap/segmap # is exactly half the size of an image and the size change on an axis is # an odd value. Then the projected value would end up being .5 # and the rounding would always round up to the next integer. If both # sides then have the same change, they are both rounded up, resulting # in more change than expected. # E.g. image height is 8, map height is 4, change is 3 at the top and 3 at # the bottom. The changes are projected to 4*(3/8) = 1.5 and both rounded # up to 2.0. Hence, the maps are changed by 4 (100% of the map height, # vs. 6 for images, which is 75% of the image height). top = _int_r(height_to * (top/height_from) - 1e-4) right = _int_r(width_to * (right/width_from) + 1e-4) bottom = _int_r(height_to * (bottom/height_from) + 1e-4) left = _int_r(width_to * (left/width_from) - 1e-4) return top, right, bottom, left def _int_r(value): return int(np.round(value)) # TODO somehow integrate this with pad() def _handle_pad_mode_param(pad_mode): pad_modes_available = { "constant", "edge", "linear_ramp", "maximum", "mean", "median", "minimum", "reflect", "symmetric", "wrap"} if pad_mode == ia.ALL: return iap.Choice(list(pad_modes_available)) if ia.is_string(pad_mode): assert pad_mode in pad_modes_available, ( "Value '%s' is not a valid pad mode. Valid pad modes are: %s." % ( pad_mode, ", ".join(pad_modes_available))) return iap.Deterministic(pad_mode) if isinstance(pad_mode, list): assert all([v in pad_modes_available for v in pad_mode]), ( "At least one in list %s is not a valid pad mode. Valid pad " "modes are: %s." % (str(pad_mode), ", ".join(pad_modes_available))) return iap.Choice(pad_mode) if isinstance(pad_mode, iap.StochasticParameter): return pad_mode raise Exception( "Expected pad_mode to be ia.ALL or string or list of strings or " "StochasticParameter, got %s." % (type(pad_mode),)) def _handle_position_parameter(position): if position == "uniform": return iap.Uniform(0.0, 1.0), iap.Uniform(0.0, 1.0) if position == "normal": return ( iap.Clip(iap.Normal(loc=0.5, scale=0.35 / 2), minval=0.0, maxval=1.0), iap.Clip(iap.Normal(loc=0.5, scale=0.35 / 2), minval=0.0, maxval=1.0) ) if position == "center": return iap.Deterministic(0.5), iap.Deterministic(0.5) if (ia.is_string(position) and re.match(r"^(left|center|right)-(top|center|bottom)$", position)): mapping = {"top": 0.0, "center": 0.5, "bottom": 1.0, "left": 0.0, "right": 1.0} return ( iap.Deterministic(mapping[position.split("-")[0]]), iap.Deterministic(mapping[position.split("-")[1]]) ) if isinstance(position, iap.StochasticParameter): return position if isinstance(position, tuple): assert len(position) == 2, ( "Expected tuple with two entries as position parameter. " "Got %d entries with types %s.." % ( len(position), str([type(item) for item in position]))) for item in position: if ia.is_single_number(item) and (item < 0 or item > 1.0): raise Exception( "Both position values must be within the value range " "[0.0, 1.0]. Got type %s with value %.8f." % ( type(item), item,)) position = [iap.Deterministic(item) if ia.is_single_number(item) else item for item in position] only_sparams = all([isinstance(item, iap.StochasticParameter) for item in position]) assert only_sparams, ( "Expected tuple with two entries that are both either " "StochasticParameter or float/int. Got types %s." % ( str([type(item) for item in position]) )) return tuple(position) raise Exception( "Expected one of the following as position parameter: string " "'uniform', string 'normal', string 'center', a string matching " "regex ^(left|center|right)-(top|center|bottom)$, a single " "StochasticParameter or a tuple of two entries, both being either " "StochasticParameter or floats or int. Got instead type %s with " "content '%s'." % ( type(position), (str(position) if len(str(position)) < 20 else str(position)[0:20] + "...") ) ) # TODO this is the same as in imgaug.py, make DRY # Added in 0.4.0. def _assert_two_or_three_dims(shape): if hasattr(shape, "shape"): shape = shape.shape assert len(shape) in [2, 3], ( "Expected image with two or three dimensions, but got %d dimensions " "and shape %s." % (len(shape), shape)) def pad(arr, top=0, right=0, bottom=0, left=0, mode="constant", cval=0): """Pad an image-like array on its top/right/bottom/left side. This function is a wrapper around :func:`numpy.pad`. Added in 0.4.0. (Previously named ``imgaug.imgaug.pad()``.) **Supported dtypes**: * ``uint8``: yes; fully tested (1) * ``uint16``: yes; fully tested (1) * ``uint32``: yes; fully tested (2) (3) * ``uint64``: yes; fully tested (2) (3) * ``int8``: yes; fully tested (1) * ``int16``: yes; fully tested (1) * ``int32``: yes; fully tested (1) * ``int64``: yes; fully tested (2) (3) * ``float16``: yes; fully tested (2) (3) * ``float32``: yes; fully tested (1) * ``float64``: yes; fully tested (1) * ``float128``: yes; fully tested (2) (3) * ``bool``: yes; tested (2) (3) - (1) Uses ``cv2`` if `mode` is one of: ``"constant"``, ``"edge"``, ``"reflect"``, ``"symmetric"``. Otherwise uses ``numpy``. - (2) Uses ``numpy``. - (3) Rejected by ``cv2``. Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray Image-like array to pad. top : int, optional Amount of pixels to add to the top side of the image. Must be ``0`` or greater. right : int, optional Amount of pixels to add to the right side of the image. Must be ``0`` or greater. bottom : int, optional Amount of pixels to add to the bottom side of the image. Must be ``0`` or greater. left : int, optional Amount of pixels to add to the left side of the image. Must be ``0`` or greater. mode : str, optional Padding mode to use. See :func:`numpy.pad` for details. In case of mode ``constant``, the parameter `cval` will be used as the ``constant_values`` parameter to :func:`numpy.pad`. In case of mode ``linear_ramp``, the parameter `cval` will be used as the ``end_values`` parameter to :func:`numpy.pad`. cval : number or iterable of number, optional Value to use for padding if `mode` is ``constant``. See :func:`numpy.pad` for details. The cval is expected to match the input array's dtype and value range. If an iterable is used, it is expected to contain one value per channel. The number of values and number of channels are expected to match. Returns ------- (H',W') ndarray or (H',W',C) ndarray Padded array with height ``H'=H+top+bottom`` and width ``W'=W+left+right``. """ import imgaug.dtypes as iadt _assert_two_or_three_dims(arr) assert all([v >= 0 for v in [top, right, bottom, left]]), ( "Expected padding amounts that are >=0, but got %d, %d, %d, %d " "(top, right, bottom, left)" % (top, right, bottom, left)) is_multi_cval = ia.is_iterable(cval) if top > 0 or right > 0 or bottom > 0 or left > 0: min_value, _, max_value = iadt.get_value_range_of_dtype(arr.dtype) # without the if here there are crashes for float128, e.g. if # cval is an int (just using float(cval) seems to not be accurate # enough) if arr.dtype.name == "float128": cval = np.float128(cval) # pylint: disable=no-member if is_multi_cval: cval = np.clip(cval, min_value, max_value) else: cval = max(min(cval, max_value), min_value) # Note that copyMakeBorder() hangs/runs endlessly if arr has an # axis of size 0 and mode is "reflect". # Numpy also complains in these cases if mode is not "constant". has_zero_sized_axis = any([axis == 0 for axis in arr.shape]) if has_zero_sized_axis: mode = "constant" mapping_mode_np_to_cv2 = { "constant": cv2.BORDER_CONSTANT, "edge": cv2.BORDER_REPLICATE, "linear_ramp": None, "maximum": None, "mean": None, "median": None, "minimum": None, "reflect": cv2.BORDER_REFLECT_101, "symmetric": cv2.BORDER_REFLECT, "wrap": None, cv2.BORDER_CONSTANT: cv2.BORDER_CONSTANT, cv2.BORDER_REPLICATE: cv2.BORDER_REPLICATE, cv2.BORDER_REFLECT_101: cv2.BORDER_REFLECT_101, cv2.BORDER_REFLECT: cv2.BORDER_REFLECT } bad_mode_cv2 = mapping_mode_np_to_cv2.get(mode, None) is None # these datatypes all simply generate a "TypeError: src data type = X # is not supported" error bad_datatype_cv2 = ( arr.dtype.name in ["uint32", "uint64", "int64", "float16", "float128", "bool"] ) # OpenCV turns the channel axis for arrays with 0 channels to 512 # TODO add direct test for this. indirectly tested via Pad bad_shape_cv2 = (arr.ndim == 3 and arr.shape[-1] == 0) if not bad_datatype_cv2 and not bad_mode_cv2 and not bad_shape_cv2: # convert cval to expected type, as otherwise we get TypeError # for np inputs kind = arr.dtype.kind if is_multi_cval: cval = [float(cval_c) if kind == "f" else int(cval_c) for cval_c in cval] else: cval = float(cval) if kind == "f" else int(cval) if arr.ndim == 2 or arr.shape[2] <= 4: # without this, only the first channel is padded with the cval, # all following channels with 0 if arr.ndim == 3 and not is_multi_cval: cval = tuple([cval] * arr.shape[2]) arr_pad = cv2.copyMakeBorder( _normalize_cv2_input_arr_(arr), top=top, bottom=bottom, left=left, right=right, borderType=mapping_mode_np_to_cv2[mode], value=cval) if arr.ndim == 3 and arr_pad.ndim == 2: arr_pad = arr_pad[..., np.newaxis] else: result = [] channel_start_idx = 0 cval = cval if is_multi_cval else tuple([cval] * arr.shape[2]) while channel_start_idx < arr.shape[2]: arr_c = arr[..., channel_start_idx:channel_start_idx+4] cval_c = cval[channel_start_idx:channel_start_idx+4] arr_pad_c = cv2.copyMakeBorder( _normalize_cv2_input_arr_(arr_c), top=top, bottom=bottom, left=left, right=right, borderType=mapping_mode_np_to_cv2[mode], value=cval_c) arr_pad_c = np.atleast_3d(arr_pad_c) result.append(arr_pad_c) channel_start_idx += 4 arr_pad = np.concatenate(result, axis=2) else: # paddings for 2d case paddings_np = [(top, bottom), (left, right)] # add paddings for 3d case if arr.ndim == 3: paddings_np.append((0, 0)) if mode == "constant": if arr.ndim > 2 and is_multi_cval: arr_pad_chans = [ np.pad(arr[..., c], paddings_np[0:2], mode=mode, constant_values=cval[c]) for c in np.arange(arr.shape[2])] arr_pad = np.stack(arr_pad_chans, axis=-1) else: arr_pad = np.pad(arr, paddings_np, mode=mode, constant_values=cval) elif mode == "linear_ramp": if arr.ndim > 2 and is_multi_cval: arr_pad_chans = [ np.pad(arr[..., c], paddings_np[0:2], mode=mode, end_values=cval[c]) for c in np.arange(arr.shape[2])] arr_pad = np.stack(arr_pad_chans, axis=-1) else: arr_pad = np.pad(arr, paddings_np, mode=mode, end_values=cval) else: arr_pad = np.pad(arr, paddings_np, mode=mode) return arr_pad return np.copy(arr) def pad_to_aspect_ratio(arr, aspect_ratio, mode="constant", cval=0, return_pad_amounts=False): """Pad an image array on its sides so that it matches a target aspect ratio. See :func:`~imgaug.imgaug.compute_paddings_for_aspect_ratio` for an explanation of how the required padding amounts are distributed per image axis. Added in 0.4.0. (Previously named ``imgaug.imgaug.pad_to_aspect_ratio()``.) **Supported dtypes**: See :func:`~imgaug.augmenters.size.pad`. Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray Image-like array to pad. aspect_ratio : float Target aspect ratio, given as width/height. E.g. ``2.0`` denotes the image having twice as much width as height. mode : str, optional Padding mode to use. See :func:`~imgaug.imgaug.pad` for details. cval : number, optional Value to use for padding if `mode` is ``constant``. See :func:`numpy.pad` for details. return_pad_amounts : bool, optional If ``False``, then only the padded image will be returned. If ``True``, a ``tuple`` with two entries will be returned, where the first entry is the padded image and the second entry are the amounts by which each image side was padded. These amounts are again a ``tuple`` of the form ``(top, right, bottom, left)``, with each value being an ``int``. Returns ------- (H',W') ndarray or (H',W',C) ndarray Padded image as ``(H',W')`` or ``(H',W',C)`` ndarray, fulfilling the given `aspect_ratio`. tuple of int Amounts by which the image was padded on each side, given as a ``tuple`` ``(top, right, bottom, left)``. This ``tuple`` is only returned if `return_pad_amounts` was set to ``True``. """ pad_top, pad_right, pad_bottom, pad_left = \ compute_paddings_to_reach_aspect_ratio(arr, aspect_ratio) arr_padded = pad( arr, top=pad_top, right=pad_right, bottom=pad_bottom, left=pad_left, mode=mode, cval=cval ) if return_pad_amounts: return arr_padded, (pad_top, pad_right, pad_bottom, pad_left) return arr_padded def pad_to_multiples_of(arr, height_multiple, width_multiple, mode="constant", cval=0, return_pad_amounts=False): """Pad an image array until its side lengths are multiples of given values. See :func:`~imgaug.imgaug.compute_paddings_for_aspect_ratio` for an explanation of how the required padding amounts are distributed per image axis. Added in 0.4.0. (Previously named ``imgaug.imgaug.pad_to_multiples_of()``.) **Supported dtypes**: See :func:`~imgaug.augmenters.size.pad`. Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray Image-like array to pad. height_multiple : None or int The desired multiple of the height. The computed padding amount will reflect a padding that increases the y axis size until it is a multiple of this value. width_multiple : None or int The desired multiple of the width. The computed padding amount will reflect a padding that increases the x axis size until it is a multiple of this value. mode : str, optional Padding mode to use. See :func:`~imgaug.imgaug.pad` for details. cval : number, optional Value to use for padding if `mode` is ``constant``. See :func:`numpy.pad` for details. return_pad_amounts : bool, optional If ``False``, then only the padded image will be returned. If ``True``, a ``tuple`` with two entries will be returned, where the first entry is the padded image and the second entry are the amounts by which each image side was padded. These amounts are again a ``tuple`` of the form ``(top, right, bottom, left)``, with each value being an integer. Returns ------- (H',W') ndarray or (H',W',C) ndarray Padded image as ``(H',W')`` or ``(H',W',C)`` ndarray. tuple of int Amounts by which the image was padded on each side, given as a ``tuple`` ``(top, right, bottom, left)``. This ``tuple`` is only returned if `return_pad_amounts` was set to ``True``. """ pad_top, pad_right, pad_bottom, pad_left = \ compute_paddings_to_reach_multiples_of( arr, height_multiple, width_multiple) arr_padded = pad( arr, top=pad_top, right=pad_right, bottom=pad_bottom, left=pad_left, mode=mode, cval=cval ) if return_pad_amounts: return arr_padded, (pad_top, pad_right, pad_bottom, pad_left) return arr_padded def compute_paddings_to_reach_aspect_ratio(arr, aspect_ratio): """Compute pad amounts required to fulfill an aspect ratio. "Pad amounts" here denotes the number of pixels that have to be added to each side to fulfill the desired constraint. The aspect ratio is given as ``ratio = width / height``. Depending on which dimension is smaller (height or width), only the corresponding sides (top/bottom or left/right) will be padded. The axis-wise padding amounts are always distributed equally over the sides of the respective axis (i.e. left and right, top and bottom). For odd pixel amounts, one pixel will be left over after the equal distribution and could be added to either side of the axis. This function will always add such a left over pixel to the bottom (y-axis) or right (x-axis) side. Added in 0.4.0. (Previously named ``imgaug.imgaug.compute_paddings_to_reach_aspect_ratio()``.) Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray or tuple of int Image-like array or shape tuple for which to compute pad amounts. aspect_ratio : float Target aspect ratio, given as width/height. E.g. ``2.0`` denotes the image having twice as much width as height. Returns ------- tuple of int Required padding amounts to reach the target aspect ratio, given as a ``tuple`` of the form ``(top, right, bottom, left)``. """ _assert_two_or_three_dims(arr) assert aspect_ratio > 0, ( "Expected to get an aspect ratio >0, got %.4f." % (aspect_ratio,)) pad_top = 0 pad_right = 0 pad_bottom = 0 pad_left = 0 shape = arr.shape if hasattr(arr, "shape") else arr height, width = shape[0:2] if height == 0: height = 1 pad_bottom += 1 if width == 0: width = 1 pad_right += 1 aspect_ratio_current = width / height if aspect_ratio_current < aspect_ratio: # image is more vertical than desired, width needs to be increased diff = (aspect_ratio * height) - width pad_right += int(np.ceil(diff / 2)) pad_left += int(np.floor(diff / 2)) elif aspect_ratio_current > aspect_ratio: # image is more horizontal than desired, height needs to be increased diff = ((1/aspect_ratio) * width) - height pad_top += int(np.floor(diff / 2)) pad_bottom += int(np.ceil(diff / 2)) return pad_top, pad_right, pad_bottom, pad_left def compute_croppings_to_reach_aspect_ratio(arr, aspect_ratio): """Compute crop amounts required to fulfill an aspect ratio. "Crop amounts" here denotes the number of pixels that have to be removed from each side to fulfill the desired constraint. The aspect ratio is given as ``ratio = width / height``. Depending on which dimension is smaller (height or width), only the corresponding sides (top/bottom or left/right) will be cropped. The axis-wise padding amounts are always distributed equally over the sides of the respective axis (i.e. left and right, top and bottom). For odd pixel amounts, one pixel will be left over after the equal distribution and could be added to either side of the axis. This function will always add such a left over pixel to the bottom (y-axis) or right (x-axis) side. If an aspect ratio cannot be reached exactly, this function will return rather one pixel too few than one pixel too many. Added in 0.4.0. Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray or tuple of int Image-like array or shape tuple for which to compute crop amounts. aspect_ratio : float Target aspect ratio, given as width/height. E.g. ``2.0`` denotes the image having twice as much width as height. Returns ------- tuple of int Required cropping amounts to reach the target aspect ratio, given as a ``tuple`` of the form ``(top, right, bottom, left)``. """ _assert_two_or_three_dims(arr) assert aspect_ratio > 0, ( "Expected to get an aspect ratio >0, got %.4f." % (aspect_ratio,)) shape = arr.shape if hasattr(arr, "shape") else arr assert shape[0] > 0, ( "Expected to get an array with height >0, got shape %s." % (shape,)) height, width = shape[0:2] aspect_ratio_current = width / height top = 0 right = 0 bottom = 0 left = 0 if aspect_ratio_current < aspect_ratio: # image is more vertical than desired, height needs to be reduced # c = H - W/r crop_amount = height - (width / aspect_ratio) crop_amount = min(crop_amount, height - 1) top = int(np.floor(crop_amount / 2)) bottom = int(np.ceil(crop_amount / 2)) elif aspect_ratio_current > aspect_ratio: # image is more horizontal than desired, width needs to be reduced # c = W - Hr crop_amount = width - height * aspect_ratio crop_amount = min(crop_amount, width - 1) left = int(np.floor(crop_amount / 2)) right = int(np.ceil(crop_amount / 2)) return top, right, bottom, left def compute_paddings_to_reach_multiples_of(arr, height_multiple, width_multiple): """Compute pad amounts until img height/width are multiples of given values. See :func:`~imgaug.imgaug.compute_paddings_for_aspect_ratio` for an explanation of how the required padding amounts are distributed per image axis. Added in 0.4.0. (Previously named ``imgaug.imgaug.compute_paddings_to_reach_multiples_of()``.) Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray or tuple of int Image-like array or shape tuple for which to compute pad amounts. height_multiple : None or int The desired multiple of the height. The computed padding amount will reflect a padding that increases the y axis size until it is a multiple of this value. width_multiple : None or int The desired multiple of the width. The computed padding amount will reflect a padding that increases the x axis size until it is a multiple of this value. Returns ------- tuple of int Required padding amounts to reach multiples of the provided values, given as a ``tuple`` of the form ``(top, right, bottom, left)``. """ def _compute_axis_value(axis_size, multiple): if multiple is None: return 0, 0 if axis_size == 0: to_pad = multiple elif axis_size % multiple == 0: to_pad = 0 else: to_pad = multiple - (axis_size % multiple) return int(np.floor(to_pad/2)), int(np.ceil(to_pad/2)) _assert_two_or_three_dims(arr) if height_multiple is not None: assert height_multiple > 0, ( "Can only pad to multiples of 1 or larger, got %d." % ( height_multiple,)) if width_multiple is not None: assert width_multiple > 0, ( "Can only pad to multiples of 1 or larger, got %d." % ( width_multiple,)) shape = arr.shape if hasattr(arr, "shape") else arr height, width = shape[0:2] top, bottom = _compute_axis_value(height, height_multiple) left, right = _compute_axis_value(width, width_multiple) return top, right, bottom, left def compute_croppings_to_reach_multiples_of(arr, height_multiple, width_multiple): """Compute croppings to reach multiples of given heights/widths. See :func:`~imgaug.imgaug.compute_paddings_for_aspect_ratio` for an explanation of how the required cropping amounts are distributed per image axis. Added in 0.4.0. Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray or tuple of int Image-like array or shape tuple for which to compute crop amounts. height_multiple : None or int The desired multiple of the height. The computed croppings will reflect a crop operation that decreases the y axis size until it is a multiple of this value. width_multiple : None or int The desired multiple of the width. The computed croppings amount will reflect a crop operation that decreases the x axis size until it is a multiple of this value. Returns ------- tuple of int Required cropping amounts to reach multiples of the provided values, given as a ``tuple`` of the form ``(top, right, bottom, left)``. """ def _compute_axis_value(axis_size, multiple): if multiple is None: return 0, 0 if axis_size == 0: to_crop = 0 elif axis_size % multiple == 0: to_crop = 0 else: to_crop = axis_size % multiple return int(np.floor(to_crop/2)), int(np.ceil(to_crop/2)) _assert_two_or_three_dims(arr) if height_multiple is not None: assert height_multiple > 0, ( "Can only crop to multiples of 1 or larger, got %d." % ( height_multiple,)) if width_multiple is not None: assert width_multiple > 0, ( "Can only crop to multiples of 1 or larger, got %d." % ( width_multiple,)) shape = arr.shape if hasattr(arr, "shape") else arr height, width = shape[0:2] top, bottom = _compute_axis_value(height, height_multiple) left, right = _compute_axis_value(width, width_multiple) return top, right, bottom, left def compute_paddings_to_reach_powers_of(arr, height_base, width_base, allow_zero_exponent=False): """Compute paddings to reach powers of given base values. For given axis size ``S``, padded size ``S'`` (``S' >= S``) and base ``B`` this function computes paddings that fulfill ``S' = B^E``, where ``E`` is any exponent from the discrete interval ``[0 .. inf)``. See :func:`~imgaug.imgaug.compute_paddings_for_aspect_ratio` for an explanation of how the required padding amounts are distributed per image axis. Added in 0.4.0. (Previously named ``imgaug.imgaug.compute_paddings_to_reach_exponents_of()``.) Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray or tuple of int Image-like array or shape tuple for which to compute pad amounts. height_base : None or int The desired base of the height. width_base : None or int The desired base of the width. allow_zero_exponent : bool, optional Whether ``E=0`` in ``S'=B^E`` is a valid value. If ``True``, axes with size ``0`` or ``1`` will be padded up to size ``B^0=1`` and axes with size ``1 < S <= B`` will be padded up to ``B^1=B``. If ``False``, the minimum output axis size is always at least ``B``. Returns ------- tuple of int Required padding amounts to fulfill ``S' = B^E`` given as a ``tuple`` of the form ``(top, right, bottom, left)``. """ def _compute_axis_value(axis_size, base): if base is None: return 0, 0 if axis_size == 0: to_pad = 1 if allow_zero_exponent else base elif axis_size <= base: to_pad = base - axis_size else: # log_{base}(axis_size) in numpy exponent = np.log(axis_size) / np.log(base) to_pad = (base ** int(np.ceil(exponent))) - axis_size return int(np.floor(to_pad/2)), int(np.ceil(to_pad/2)) _assert_two_or_three_dims(arr) if height_base is not None: assert height_base > 1, ( "Can only pad to base larger than 1, got %d." % (height_base,)) if width_base is not None: assert width_base > 1, ( "Can only pad to base larger than 1, got %d." % (width_base,)) shape = arr.shape if hasattr(arr, "shape") else arr height, width = shape[0:2] top, bottom = _compute_axis_value(height, height_base) left, right = _compute_axis_value(width, width_base) return top, right, bottom, left def compute_croppings_to_reach_powers_of(arr, height_base, width_base, allow_zero_exponent=False): """Compute croppings to reach powers of given base values. For given axis size ``S``, cropped size ``S'`` (``S' <= S``) and base ``B`` this function computes croppings that fulfill ``S' = B^E``, where ``E`` is any exponent from the discrete interval ``[0 .. inf)``. See :func:`~imgaug.imgaug.compute_paddings_for_aspect_ratio` for an explanation of how the required cropping amounts are distributed per image axis. .. note:: For axes where ``S == 0``, this function alwayws returns zeros as croppings. For axes where ``1 <= S < B`` see parameter `allow_zero_exponent`. Added in 0.4.0. Parameters ---------- arr : (H,W) ndarray or (H,W,C) ndarray or tuple of int Image-like array or shape tuple for which to compute crop amounts. height_base : None or int The desired base of the height. width_base : None or int The desired base of the width. allow_zero_exponent : bool Whether ``E=0`` in ``S'=B^E`` is a valid value. If ``True``, axes with size ``1 <= S < B`` will be cropped to size ``B^0=1``. If ``False``, axes with sizes ``S < B`` will not be changed. Returns ------- tuple of int Required cropping amounts to fulfill ``S' = B^E`` given as a ``tuple`` of the form ``(top, right, bottom, left)``. """ def _compute_axis_value(axis_size, base): if base is None: return 0, 0 if axis_size == 0: to_crop = 0 elif axis_size < base: # crop down to B^0 = 1 to_crop = axis_size - 1 if allow_zero_exponent else 0 else: # log_{base}(axis_size) in numpy exponent = np.log(axis_size) / np.log(base) to_crop = axis_size - (base ** int(exponent)) return int(np.floor(to_crop/2)), int(np.ceil(to_crop/2)) _assert_two_or_three_dims(arr) if height_base is not None: assert height_base > 1, ( "Can only crop to base larger than 1, got %d." % (height_base,)) if width_base is not None: assert width_base > 1, ( "Can only crop to base larger than 1, got %d." % (width_base,)) shape = arr.shape if hasattr(arr, "shape") else arr height, width = shape[0:2] top, bottom = _compute_axis_value(height, height_base) left, right = _compute_axis_value(width, width_base) return top, right, bottom, left @ia.deprecated(alt_func="Resize", comment="Resize has the exactly same interface as Scale.") def Scale(*args, **kwargs): """Augmenter that resizes images to specified heights and widths.""" # pylint: disable=invalid-name return Resize(*args, **kwargs) class Resize(meta.Augmenter): """Augmenter that resizes images to specified heights and widths. **Supported dtypes**: See :func:`~imgaug.imgaug.imresize_many_images`. Parameters ---------- size : 'keep' or int or float or tuple of int or tuple of float or list of int or list of float or imgaug.parameters.StochasticParameter or dict The new size of the images. * If this has the string value ``keep``, the original height and width values will be kept (image is not resized). * If this is an ``int``, this value will always be used as the new height and width of the images. * If this is a ``float`` ``v``, then per image the image's height ``H`` and width ``W`` will be changed to ``H*v`` and ``W*v``. * If this is a ``tuple``, it is expected to have two entries ``(a, b)``. If at least one of these are ``float`` s, a value will be sampled from range ``[a, b]`` and used as the ``float`` value to resize the image (see above). If both are ``int`` s, a value will be sampled from the discrete range ``[a..b]`` and used as the integer value to resize the image (see above). * If this is a ``list``, a random value from the ``list`` will be picked to resize the image. All values in the ``list`` must be ``int`` s or ``float`` s (no mixture is possible). * If this is a ``StochasticParameter``, then this parameter will first be queried once per image. The resulting value will be used for both height and width. * If this is a ``dict``, it may contain the keys ``height`` and ``width`` or the keys ``shorter-side`` and ``longer-side``. Each key may have the same datatypes as above and describes the scaling on x and y-axis or the shorter and longer axis, respectively. Both axis are sampled independently. Additionally, one of the keys may have the value ``keep-aspect-ratio``, which means that the respective side of the image will be resized so that the original aspect ratio is kept. This is useful when only resizing one image size by a pixel value (e.g. resize images to a height of ``64`` pixels and resize the width so that the overall aspect ratio is maintained). interpolation : imgaug.ALL or int or str or list of int or list of str or imgaug.parameters.StochasticParameter, optional Interpolation to use. * If ``imgaug.ALL``, then a random interpolation from ``nearest``, ``linear``, ``area`` or ``cubic`` will be picked (per image). * If ``int``, then this interpolation will always be used. Expected to be any of the following: ``cv2.INTER_NEAREST``, ``cv2.INTER_LINEAR``, ``cv2.INTER_AREA``, ``cv2.INTER_CUBIC`` * If string, then this interpolation will always be used. Expected to be any of the following: ``nearest``, ``linear``, ``area``, ``cubic`` * If ``list`` of ``int`` / ``str``, then a random one of the values will be picked per image as the interpolation. * If a ``StochasticParameter``, then this parameter will be queried per image and is expected to return an ``int`` or ``str``. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.Resize(32) Resize all images to ``32x32`` pixels. >>> aug = iaa.Resize(0.5) Resize all images to ``50`` percent of their original size. >>> aug = iaa.Resize((16, 22)) Resize all images to a random height and width within the discrete interval ``[16..22]`` (uniformly sampled per image). >>> aug = iaa.Resize((0.5, 0.75)) Resize all any input image so that its height (``H``) and width (``W``) become ``H*v`` and ``W*v``, where ``v`` is uniformly sampled from the interval ``[0.5, 0.75]``. >>> aug = iaa.Resize([16, 32, 64]) Resize all images either to ``16x16``, ``32x32`` or ``64x64`` pixels. >>> aug = iaa.Resize({"height": 32}) Resize all images to a height of ``32`` pixels and keeps the original width. >>> aug = iaa.Resize({"height": 32, "width": 48}) Resize all images to a height of ``32`` pixels and a width of ``48``. >>> aug = iaa.Resize({"height": 32, "width": "keep-aspect-ratio"}) Resize all images to a height of ``32`` pixels and resizes the x-axis (width) so that the aspect ratio is maintained. >>> aug = iaa.Resize( >>> {"shorter-side": 224, "longer-side": "keep-aspect-ratio"}) Resize all images to a height/width of ``224`` pixels, depending on which axis is shorter and resize the other axis so that the aspect ratio is maintained. >>> aug = iaa.Resize({"height": (0.5, 0.75), "width": [16, 32, 64]}) Resize all images to a height of ``H*v``, where ``H`` is the original height and ``v`` is a random value sampled from the interval ``[0.5, 0.75]``. The width/x-axis of each image is resized to either ``16`` or ``32`` or ``64`` pixels. >>> aug = iaa.Resize(32, interpolation=["linear", "cubic"]) Resize all images to ``32x32`` pixels. Randomly use either ``linear`` or ``cubic`` interpolation. """ def __init__(self, size, interpolation="cubic", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(Resize, self).__init__( seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.size, self.size_order = self._handle_size_arg(size, False) self.interpolation = self._handle_interpolation_arg(interpolation) @classmethod def _handle_size_arg(cls, size, subcall): def _dict_to_size_tuple(val1, val2): kaa = "keep-aspect-ratio" not_both_kaa = (val1 != kaa or val2 != kaa) assert not_both_kaa, ( "Expected at least one value to not be \"keep-aspect-ratio\", " "but got it two times.") size_tuple = [] for k in [val1, val2]: if k in ["keep-aspect-ratio", "keep"]: entry = iap.Deterministic(k) else: entry = cls._handle_size_arg(k, True) size_tuple.append(entry) return tuple(size_tuple) def _contains_any_key(dict_, keys): return any([key in dict_ for key in keys]) # HW = height, width # SL = shorter, longer size_order = "HW" if size == "keep": result = iap.Deterministic("keep") elif ia.is_single_number(size): assert size > 0, "Expected only values > 0, got %s" % (size,) result = iap.Deterministic(size) elif not subcall and isinstance(size, dict): if len(size.keys()) == 0: result = iap.Deterministic("keep") elif _contains_any_key(size, ["height", "width"]): height = size.get("height", "keep") width = size.get("width", "keep") result = _dict_to_size_tuple(height, width) elif _contains_any_key(size, ["shorter-side", "longer-side"]): shorter = size.get("shorter-side", "keep") longer = size.get("longer-side", "keep") result = _dict_to_size_tuple(shorter, longer) size_order = "SL" else: raise ValueError( "Expected dictionary containing no keys, " "the keys \"height\" and/or \"width\", " "or the keys \"shorter-side\" and/or \"longer-side\". " "Got keys: %s." % (str(size.keys()),)) elif isinstance(size, tuple): assert len(size) == 2, ( "Expected size tuple to contain exactly 2 values, " "got %d." % (len(size),)) assert size[0] > 0 and size[1] > 0, ( "Expected size tuple to only contain values >0, " "got %d and %d." % (size[0], size[1])) if ia.is_single_float(size[0]) or ia.is_single_float(size[1]): result = iap.Uniform(size[0], size[1]) else: result = iap.DiscreteUniform(size[0], size[1]) elif isinstance(size, list): if len(size) == 0: result = iap.Deterministic("keep") else: all_int = all([ia.is_single_integer(v) for v in size]) all_float = all([ia.is_single_float(v) for v in size]) assert all_int or all_float, ( "Expected to get only integers or floats.") assert all([v > 0 for v in size]), ( "Expected all values to be >0.") result = iap.Choice(size) elif isinstance(size, iap.StochasticParameter): result = size else: raise ValueError( "Expected number, tuple of two numbers, list of numbers, " "dictionary of form " "{'height': number/tuple/list/'keep-aspect-ratio'/'keep', " "'width': }, dictionary of form " "{'shorter-side': number/tuple/list/'keep-aspect-ratio'/" "'keep', 'longer-side': } " "or StochasticParameter, got %s." % (type(size),) ) if subcall: return result return result, size_order @classmethod def _handle_interpolation_arg(cls, interpolation): if interpolation == ia.ALL: interpolation = iap.Choice( ["nearest", "linear", "area", "cubic"]) elif ia.is_single_integer(interpolation): interpolation = iap.Deterministic(interpolation) elif ia.is_string(interpolation): interpolation = iap.Deterministic(interpolation) elif ia.is_iterable(interpolation): interpolation = iap.Choice(interpolation) elif isinstance(interpolation, iap.StochasticParameter): pass else: raise Exception( "Expected int or string or iterable or StochasticParameter, " "got %s." % (type(interpolation),)) return interpolation # Added in 0.4.0. def _augment_batch_(self, batch, random_state, parents, hooks): nb_rows = batch.nb_rows samples = self._draw_samples(nb_rows, random_state) if batch.images is not None: batch.images = self._augment_images_by_samples(batch.images, samples) if batch.heatmaps is not None: # TODO this uses the same interpolation as for images for heatmaps # while other augmenters resort to cubic batch.heatmaps = self._augment_maps_by_samples( batch.heatmaps, "arr_0to1", samples) if batch.segmentation_maps is not None: batch.segmentation_maps = self._augment_maps_by_samples( batch.segmentation_maps, "arr", (samples[0], samples[1], [None] * nb_rows)) for augm_name in ["keypoints", "bounding_boxes", "polygons", "line_strings"]: augm_value = getattr(batch, augm_name) if augm_value is not None: func = functools.partial( self._augment_keypoints_by_samples, samples=samples) cbaois = self._apply_to_cbaois_as_keypoints(augm_value, func) setattr(batch, augm_name, cbaois) return batch # Added in 0.4.0. def _augment_images_by_samples(self, images, samples): input_was_array = False input_dtype = None if ia.is_np_array(images): input_was_array = True input_dtype = images.dtype samples_a, samples_b, samples_ip = samples result = [] for i, image in enumerate(images): h, w = self._compute_height_width(image.shape, samples_a[i], samples_b[i], self.size_order) image_rs = ia.imresize_single_image(image, (h, w), interpolation=samples_ip[i]) result.append(image_rs) if input_was_array: all_same_size = (len({image.shape for image in result}) == 1) if all_same_size: result = np.array(result, dtype=input_dtype) return result # Added in 0.4.0. def _augment_maps_by_samples(self, augmentables, arr_attr_name, samples): result = [] samples_h, samples_w, samples_ip = samples for i, augmentable in enumerate(augmentables): arr = getattr(augmentable, arr_attr_name) arr_shape = arr.shape img_shape = augmentable.shape h_img, w_img = self._compute_height_width( img_shape, samples_h[i], samples_w[i], self.size_order) h = int(np.round(h_img * (arr_shape[0] / img_shape[0]))) w = int(np.round(w_img * (arr_shape[1] / img_shape[1]))) h = max(h, 1) w = max(w, 1) if samples_ip[0] is not None: # TODO change this for heatmaps to always have cubic or # automatic interpolation? augmentable_resize = augmentable.resize( (h, w), interpolation=samples_ip[i]) else: augmentable_resize = augmentable.resize((h, w)) augmentable_resize.shape = (h_img, w_img) + img_shape[2:] result.append(augmentable_resize) return result # Added in 0.4.0. def _augment_keypoints_by_samples(self, kpsois, samples): result = [] samples_a, samples_b, _samples_ip = samples for i, kpsoi in enumerate(kpsois): h, w = self._compute_height_width( kpsoi.shape, samples_a[i], samples_b[i], self.size_order) new_shape = (h, w) + kpsoi.shape[2:] keypoints_on_image_rs = kpsoi.on_(new_shape) result.append(keypoints_on_image_rs) return result def _draw_samples(self, nb_images, random_state): rngs = random_state.duplicate(3) if isinstance(self.size, tuple): samples_h = self.size[0].draw_samples(nb_images, random_state=rngs[0]) samples_w = self.size[1].draw_samples(nb_images, random_state=rngs[1]) else: samples_h = self.size.draw_samples(nb_images, random_state=rngs[0]) samples_w = samples_h samples_ip = self.interpolation.draw_samples(nb_images, random_state=rngs[2]) return samples_h, samples_w, samples_ip @classmethod def _compute_height_width(cls, image_shape, sample_a, sample_b, size_order): imh, imw = image_shape[0:2] if size_order == 'SL': # size order: short, long if imh < imw: h, w = sample_a, sample_b else: w, h = sample_a, sample_b else: # size order: height, width h, w = sample_a, sample_b if ia.is_single_float(h): assert h > 0, "Expected 'h' to be >0, got %.4f" % (h,) h = int(np.round(imh * h)) h = h if h > 0 else 1 elif h == "keep": h = imh if ia.is_single_float(w): assert w > 0, "Expected 'w' to be >0, got %.4f" % (w,) w = int(np.round(imw * w)) w = w if w > 0 else 1 elif w == "keep": w = imw # at least the checks for keep-aspect-ratio must come after # the float checks, as they are dependent on the results # this is also why these are not written as elifs if h == "keep-aspect-ratio": h_per_w_orig = imh / imw h = int(np.round(w * h_per_w_orig)) if w == "keep-aspect-ratio": w_per_h_orig = imw / imh w = int(np.round(h * w_per_h_orig)) return h, w def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.size, self.interpolation, self.size_order] class _CropAndPadSamplingResult(object): def __init__(self, crop_top, crop_right, crop_bottom, crop_left, pad_top, pad_right, pad_bottom, pad_left, pad_mode, pad_cval): self.crop_top = crop_top self.crop_right = crop_right self.crop_bottom = crop_bottom self.crop_left = crop_left self.pad_top = pad_top self.pad_right = pad_right self.pad_bottom = pad_bottom self.pad_left = pad_left self.pad_mode = pad_mode self.pad_cval = pad_cval @property def croppings(self): """Get absolute pixel amounts of croppings as a TRBL tuple.""" return self.crop_top, self.crop_right, self.crop_bottom, self.crop_left @property def paddings(self): """Get absolute pixel amounts of paddings as a TRBL tuple.""" return self.pad_top, self.pad_right, self.pad_bottom, self.pad_left class CropAndPad(meta.Augmenter): """Crop/pad images by pixel amounts or fractions of image sizes. Cropping removes pixels at the sides (i.e. extracts a subimage from a given full image). Padding adds pixels to the sides (e.g. black pixels). This augmenter will never crop images below a height or width of ``1``. .. note:: This augmenter automatically resizes images back to their original size after it has augmented them. To deactivate this, add the parameter ``keep_size=False``. **Supported dtypes**: if (keep_size=False): * ``uint8``: yes; fully tested * ``uint16``: yes; tested * ``uint32``: yes; tested * ``uint64``: yes; tested * ``int8``: yes; tested * ``int16``: yes; tested * ``int32``: yes; tested * ``int64``: yes; tested * ``float16``: yes; tested * ``float32``: yes; tested * ``float64``: yes; tested * ``float128``: yes; tested * ``bool``: yes; tested if (keep_size=True): minimum of ( ``imgaug.augmenters.size.CropAndPad(keep_size=False)``, :func:`~imgaug.imgaug.imresize_many_images` ) Parameters ---------- px : None or int or imgaug.parameters.StochasticParameter or tuple, optional The number of pixels to crop (negative values) or pad (positive values) on each side of the image. Either this or the parameter `percent` may be set, not both at the same time. * If ``None``, then pixel-based cropping/padding will not be used. * If ``int``, then that exact number of pixels will always be cropped/padded. * If ``StochasticParameter``, then that parameter will be used for each image. Four samples will be drawn per image (top, right, bottom, left), unless `sample_independently` is set to ``False``, as then only one value will be sampled per image and used for all sides. * If a ``tuple`` of two ``int`` s with values ``a`` and ``b``, then each side will be cropped/padded by a random amount sampled uniformly per image and side from the inteval ``[a, b]``. If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of four entries, then the entries represent top, right, bottom, left. Each entry may be a single ``int`` (always crop/pad by exactly that value), a ``tuple`` of two ``int`` s ``a`` and ``b`` (crop/pad by an amount within ``[a, b]``), a ``list`` of ``int`` s (crop/pad by a random value that is contained in the ``list``) or a ``StochasticParameter`` (sample the amount to crop/pad from that parameter). percent : None or number or imgaug.parameters.StochasticParameter or tuple, optional The number of pixels to crop (negative values) or pad (positive values) on each side of the image given as a *fraction* of the image height/width. E.g. if this is set to ``-0.1``, the augmenter will always crop away ``10%`` of the image's height at both the top and the bottom (both ``10%`` each), as well as ``10%`` of the width at the right and left. Expected value range is ``(-1.0, inf)``. Either this or the parameter `px` may be set, not both at the same time. * If ``None``, then fraction-based cropping/padding will not be used. * If ``number``, then that fraction will always be cropped/padded. * If ``StochasticParameter``, then that parameter will be used for each image. Four samples will be drawn per image (top, right, bottom, left). If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of two ``float`` s with values ``a`` and ``b``, then each side will be cropped/padded by a random fraction sampled uniformly per image and side from the interval ``[a, b]``. If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of four entries, then the entries represent top, right, bottom, left. Each entry may be a single ``float`` (always crop/pad by exactly that percent value), a ``tuple`` of two ``float`` s ``a`` and ``b`` (crop/pad by a fraction from ``[a, b]``), a ``list`` of ``float`` s (crop/pad by a random value that is contained in the list) or a ``StochasticParameter`` (sample the percentage to crop/pad from that parameter). pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional Padding mode to use. The available modes match the numpy padding modes, i.e. ``constant``, ``edge``, ``linear_ramp``, ``maximum``, ``median``, ``minimum``, ``reflect``, ``symmetric``, ``wrap``. The modes ``constant`` and ``linear_ramp`` use extra values, which are provided by ``pad_cval`` when necessary. See :func:`~imgaug.imgaug.pad` for more details. * If ``imgaug.ALL``, then a random mode from all available modes will be sampled per image. * If a ``str``, it will be used as the pad mode for all images. * If a ``list`` of ``str``, a random one of these will be sampled per image and used as the mode. * If ``StochasticParameter``, a random mode will be sampled from this parameter per image. pad_cval : number or tuple of number list of number or imgaug.parameters.StochasticParameter, optional The constant value to use if the pad mode is ``constant`` or the end value to use if the mode is ``linear_ramp``. See :func:`~imgaug.imgaug.pad` for more details. * If ``number``, then that value will be used. * If a ``tuple`` of two ``number`` s and at least one of them is a ``float``, then a random number will be uniformly sampled per image from the continuous interval ``[a, b]`` and used as the value. If both ``number`` s are ``int`` s, the interval is discrete. * If a ``list`` of ``number``, then a random value will be chosen from the elements of the ``list`` and used as the value. * If ``StochasticParameter``, a random value will be sampled from that parameter per image. keep_size : bool, optional After cropping and padding, the result image will usually have a different height/width compared to the original input image. If this parameter is set to ``True``, then the cropped/padded image will be resized to the input image's size, i.e. the augmenter's output shape is always identical to the input shape. sample_independently : bool, optional If ``False`` *and* the values for `px`/`percent` result in exactly *one* probability distribution for all image sides, only one single value will be sampled from that probability distribution and used for all sides. I.e. the crop/pad amount then is the same for all sides. If ``True``, four values will be sampled independently, one per side. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CropAndPad(px=(-10, 0)) Crop each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[-10..0]``. >>> aug = iaa.CropAndPad(px=(0, 10)) Pad each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. The padding happens by zero-padding, i.e. it adds black pixels (default setting). >>> aug = iaa.CropAndPad(px=(0, 10), pad_mode="edge") Pad each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. The padding uses the ``edge`` mode from numpy's pad function, i.e. the pixel colors around the image sides are repeated. >>> aug = iaa.CropAndPad(px=(0, 10), pad_mode=["constant", "edge"]) Similar to the previous example, but uses zero-padding (``constant``) for half of the images and ``edge`` padding for the other half. >>> aug = iaa.CropAndPad(px=(0, 10), pad_mode=ia.ALL, pad_cval=(0, 255)) Similar to the previous example, but uses any available padding mode. In case the padding mode ends up being ``constant`` or ``linear_ramp``, and random intensity is uniformly sampled (once per image) from the discrete interval ``[0..255]`` and used as the intensity of the new pixels. >>> aug = iaa.CropAndPad(px=(0, 10), sample_independently=False) Pad each side by a random pixel value sampled uniformly once per image from the discrete interval ``[0..10]``. Each sampled value is used for *all* sides of the corresponding image. >>> aug = iaa.CropAndPad(px=(0, 10), keep_size=False) Pad each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. Afterwards, do **not** resize the padded image back to the input image's size. This will increase the image's height and width by a maximum of ``20`` pixels. >>> aug = iaa.CropAndPad(px=((0, 10), (0, 5), (0, 10), (0, 5))) Pad the top and bottom by a random pixel value sampled uniformly from the discrete interval ``[0..10]``. Pad the left and right analogously by a random value sampled from ``[0..5]``. Each value is always sampled independently. >>> aug = iaa.CropAndPad(percent=(0, 0.1)) Pad each side by a random fraction sampled uniformly from the continuous interval ``[0.0, 0.10]``. The fraction is sampled once per image and side. E.g. a sampled fraction of ``0.1`` for the top side would pad by ``0.1*H``, where ``H`` is the height of the input image. >>> aug = iaa.CropAndPad( >>> percent=([0.05, 0.1], [0.05, 0.1], [0.05, 0.1], [0.05, 0.1])) Pads each side by either ``5%`` or ``10%``. The values are sampled once per side and image. >>> aug = iaa.CropAndPad(px=(-10, 10)) Sample uniformly per image and side a value ``v`` from the discrete range ``[-10..10]``. Then either crop (negative sample) or pad (positive sample) the side by ``v`` pixels. """ def __init__(self, px=None, percent=None, pad_mode="constant", pad_cval=0, keep_size=True, sample_independently=True, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): # pylint: disable=invalid-name super(CropAndPad, self).__init__( seed=seed, name=name, random_state=random_state, deterministic=deterministic) if px is None and percent is None: percent = (-0.1, 0.1) self.mode, self.all_sides, self.top, self.right, self.bottom, \ self.left = self._handle_px_and_percent_args(px, percent) self.pad_mode = _handle_pad_mode_param(pad_mode) # TODO enable ALL here, like in e.g. Affine self.pad_cval = iap.handle_discrete_param( pad_cval, "pad_cval", value_range=None, tuple_to_uniform=True, list_to_choice=True, allow_floats=True) self.keep_size = keep_size self.sample_independently = sample_independently # set these to None to use the same values as sampled for the # images (not tested) self._pad_mode_heatmaps = "constant" self._pad_mode_segmentation_maps = "constant" self._pad_cval_heatmaps = 0.0 self._pad_cval_segmentation_maps = 0 @classmethod def _handle_px_and_percent_args(cls, px, percent): # pylint: disable=invalid-name all_sides = None top, right, bottom, left = None, None, None, None if px is None and percent is None: mode = "noop" elif px is not None and percent is not None: raise Exception("Can only pad by pixels or percent, not both.") elif px is not None: mode = "px" all_sides, top, right, bottom, left = cls._handle_px_arg(px) else: # = elif percent is not None: mode = "percent" all_sides, top, right, bottom, left = cls._handle_percent_arg( percent) return mode, all_sides, top, right, bottom, left @classmethod def _handle_px_arg(cls, px): # pylint: disable=invalid-name all_sides = None top, right, bottom, left = None, None, None, None if ia.is_single_integer(px): all_sides = iap.Deterministic(px) elif isinstance(px, tuple): assert len(px) in [2, 4], ( "Expected 'px' given as a tuple to contain 2 or 4 " "entries, got %d." % (len(px),)) def handle_param(p): if ia.is_single_integer(p): return iap.Deterministic(p) if isinstance(p, tuple): assert len(p) == 2, ( "Expected tuple of 2 values, got %d." % (len(p))) only_ints = ( ia.is_single_integer(p[0]) and ia.is_single_integer(p[1])) assert only_ints, ( "Expected tuple of integers, got %s and %s." % ( type(p[0]), type(p[1]))) return iap.DiscreteUniform(p[0], p[1]) if isinstance(p, list): assert len(p) > 0, ( "Expected non-empty list, but got empty one.") assert all([ia.is_single_integer(val) for val in p]), ( "Expected list of ints, got types %s." % ( ", ".join([str(type(v)) for v in p]))) return iap.Choice(p) if isinstance(p, iap.StochasticParameter): return p raise Exception( "Expected int, tuple of two ints, list of ints or " "StochasticParameter, got type %s." % (type(p),)) if len(px) == 2: all_sides = handle_param(px) else: # len == 4 top = handle_param(px[0]) right = handle_param(px[1]) bottom = handle_param(px[2]) left = handle_param(px[3]) elif isinstance(px, iap.StochasticParameter): top = right = bottom = left = px else: raise Exception( "Expected int, tuple of 4 " "ints/tuples/lists/StochasticParameters or " "StochasticParameter, got type %s." % (type(px),)) return all_sides, top, right, bottom, left @classmethod def _handle_percent_arg(cls, percent): all_sides = None top, right, bottom, left = None, None, None, None if ia.is_single_number(percent): assert percent > -1.0, ( "Expected 'percent' to be >-1.0, got %.4f." % (percent,)) all_sides = iap.Deterministic(percent) elif isinstance(percent, tuple): assert len(percent) in [2, 4], ( "Expected 'percent' given as a tuple to contain 2 or 4 " "entries, got %d." % (len(percent),)) def handle_param(p): if ia.is_single_number(p): return iap.Deterministic(p) if isinstance(p, tuple): assert len(p) == 2, ( "Expected tuple of 2 values, got %d." % (len(p),)) only_numbers = ( ia.is_single_number(p[0]) and ia.is_single_number(p[1])) assert only_numbers, ( "Expected tuple of numbers, got %s and %s." % ( type(p[0]), type(p[1]))) assert p[0] > -1.0 and p[1] > -1.0, ( "Expected tuple of values >-1.0, got %.4f and " "%.4f." % (p[0], p[1])) return iap.Uniform(p[0], p[1]) if isinstance(p, list): assert len(p) > 0, ( "Expected non-empty list, but got empty one.") assert all([ia.is_single_number(val) for val in p]), ( "Expected list of numbers, got types %s." % ( ", ".join([str(type(v)) for v in p]))) assert all([val > -1.0 for val in p]), ( "Expected list of values >-1.0, got values %s." % ( ", ".join(["%.4f" % (v,) for v in p]))) return iap.Choice(p) if isinstance(p, iap.StochasticParameter): return p raise Exception( "Expected int, tuple of two ints, list of ints or " "StochasticParameter, got type %s." % (type(p),)) if len(percent) == 2: all_sides = handle_param(percent) else: # len == 4 top = handle_param(percent[0]) right = handle_param(percent[1]) bottom = handle_param(percent[2]) left = handle_param(percent[3]) elif isinstance(percent, iap.StochasticParameter): top = right = bottom = left = percent else: raise Exception( "Expected number, tuple of 4 " "numbers/tuples/lists/StochasticParameters or " "StochasticParameter, got type %s." % (type(percent),)) return all_sides, top, right, bottom, left # Added in 0.4.0. def _augment_batch_(self, batch, random_state, parents, hooks): shapes = batch.get_rowwise_shapes() samples = self._draw_samples(random_state, shapes) if batch.images is not None: batch.images = self._augment_images_by_samples(batch.images, samples) if batch.heatmaps is not None: batch.heatmaps = self._augment_maps_by_samples( batch.heatmaps, self._pad_mode_heatmaps, self._pad_cval_heatmaps, samples) if batch.segmentation_maps is not None: batch.segmentation_maps = self._augment_maps_by_samples( batch.segmentation_maps, self._pad_mode_segmentation_maps, self._pad_cval_segmentation_maps, samples) for augm_name in ["keypoints", "bounding_boxes", "polygons", "line_strings"]: augm_value = getattr(batch, augm_name) if augm_value is not None: func = functools.partial( self._augment_keypoints_by_samples, samples=samples) cbaois = self._apply_to_cbaois_as_keypoints(augm_value, func) setattr(batch, augm_name, cbaois) return batch # Added in 0.4.0. def _augment_images_by_samples(self, images, samples): result = [] for i, image in enumerate(images): samples_i = samples[i] image_cr_pa = _crop_and_pad_arr( image, samples_i.croppings, samples_i.paddings, samples_i.pad_mode, samples_i.pad_cval, self.keep_size) result.append(image_cr_pa) if ia.is_np_array(images): if self.keep_size: result = np.array(result, dtype=images.dtype) else: nb_shapes = len({image.shape for image in result}) if nb_shapes == 1: result = np.array(result, dtype=images.dtype) return result # Added in 0.4.0. def _augment_maps_by_samples(self, augmentables, pad_mode, pad_cval, samples): result = [] for i, augmentable in enumerate(augmentables): samples_img = samples[i] augmentable = _crop_and_pad_hms_or_segmaps_( augmentable, croppings_img=samples_img.croppings, paddings_img=samples_img.paddings, pad_mode=(pad_mode if pad_mode is not None else samples_img.pad_mode), pad_cval=(pad_cval if pad_cval is not None else samples_img.pad_cval), keep_size=self.keep_size ) result.append(augmentable) return result # Added in 0.4.0. def _augment_keypoints_by_samples(self, keypoints_on_images, samples): result = [] for i, keypoints_on_image in enumerate(keypoints_on_images): samples_i = samples[i] kpsoi_aug = _crop_and_pad_kpsoi_( keypoints_on_image, croppings_img=samples_i.croppings, paddings_img=samples_i.paddings, keep_size=self.keep_size) result.append(kpsoi_aug) return result def _draw_samples(self, random_state, shapes): nb_rows = len(shapes) if self.mode == "noop": top = right = bottom = left = np.full((nb_rows,), 0, dtype=np.int32) else: if self.all_sides is not None: if self.sample_independently: samples = self.all_sides.draw_samples( (nb_rows, 4), random_state=random_state) top = samples[:, 0] right = samples[:, 1] bottom = samples[:, 2] left = samples[:, 3] else: sample = self.all_sides.draw_samples( (nb_rows,), random_state=random_state) top = right = bottom = left = sample else: top = self.top.draw_samples( (nb_rows,), random_state=random_state) right = self.right.draw_samples( (nb_rows,), random_state=random_state) bottom = self.bottom.draw_samples( (nb_rows,), random_state=random_state) left = self.left.draw_samples( (nb_rows,), random_state=random_state) if self.mode == "px": # no change necessary for pixel values pass elif self.mode == "percent": # percentage values have to be transformed to pixel values shapes_arr = np.array([shape[0:2] for shape in shapes], dtype=np.float32) heights = shapes_arr[:, 0] widths = shapes_arr[:, 1] top = np.round(heights * top).astype(np.int32) right = np.round(widths * right).astype(np.int32) bottom = np.round(heights * bottom).astype(np.int32) left = np.round(widths * left).astype(np.int32) else: raise Exception("Invalid mode") def _only_above_zero(arr): arr = np.copy(arr) mask = (arr < 0) arr[mask] = 0 return arr crop_top = _only_above_zero((-1) * top) crop_right = _only_above_zero((-1) * right) crop_bottom = _only_above_zero((-1) * bottom) crop_left = _only_above_zero((-1) * left) pad_top = _only_above_zero(top) pad_right = _only_above_zero(right) pad_bottom = _only_above_zero(bottom) pad_left = _only_above_zero(left) pad_mode = self.pad_mode.draw_samples((nb_rows,), random_state=random_state) pad_cval = self.pad_cval.draw_samples((nb_rows,), random_state=random_state) # TODO vectorize this part -- especially return only one instance result = [] for i, shape in enumerate(shapes): height, width = shape[0:2] crop_top_i, crop_right_i, crop_bottom_i, crop_left_i = \ _crop_prevent_zero_size( height, width, crop_top[i], crop_right[i], crop_bottom[i], crop_left[i]) # add here any_crop_y to not warn in case of zero height/width # images any_crop_y = (crop_top_i > 0 or crop_bottom_i > 0) if any_crop_y and crop_top_i + crop_bottom_i >= height: ia.warn( "Expected generated crop amounts in CropAndPad for top and " "bottom image side to be less than the image's height, but " "got %d (top) and %d (bottom) vs. image height %d. This " "will result in an image with output height=1 (if input " "height was >=1) or output height=0 (if input height " "was 0)." % (crop_top_i, crop_bottom_i, height)) # add here any_crop_x to not warn in case of zero height/width # images any_crop_x = (crop_left_i > 0 or crop_right_i > 0) if any_crop_x and crop_left_i + crop_right_i >= width: ia.warn( "Expected generated crop amounts in CropAndPad for left " "and right image side to be less than the image's width, " "but got %d (left) and %d (right) vs. image width %d. " "This will result in an image with output width=1 (if " "input width was >=1) or output width=0 (if input width " "was 0)." % (crop_left_i, crop_right_i, width)) result.append( _CropAndPadSamplingResult( crop_top=crop_top_i, crop_right=crop_right_i, crop_bottom=crop_bottom_i, crop_left=crop_left_i, pad_top=pad_top[i], pad_right=pad_right[i], pad_bottom=pad_bottom[i], pad_left=pad_left[i], pad_mode=pad_mode[i], pad_cval=pad_cval[i])) return result def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.all_sides, self.top, self.right, self.bottom, self.left, self.pad_mode, self.pad_cval] class Pad(CropAndPad): """Pad images, i.e. adds columns/rows of pixels to them. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropAndPad`. Parameters ---------- px : None or int or imgaug.parameters.StochasticParameter or tuple, optional The number of pixels to pad on each side of the image. Expected value range is ``[0, inf)``. Either this or the parameter `percent` may be set, not both at the same time. * If ``None``, then pixel-based padding will not be used. * If ``int``, then that exact number of pixels will always be padded. * If ``StochasticParameter``, then that parameter will be used for each image. Four samples will be drawn per image (top, right, bottom, left), unless `sample_independently` is set to ``False``, as then only one value will be sampled per image and used for all sides. * If a ``tuple`` of two ``int`` s with values ``a`` and ``b``, then each side will be padded by a random amount sampled uniformly per image and side from the inteval ``[a, b]``. If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of four entries, then the entries represent top, right, bottom, left. Each entry may be a single ``int`` (always pad by exactly that value), a ``tuple`` of two ``int`` s ``a`` and ``b`` (pad by an amount within ``[a, b]``), a ``list`` of ``int`` s (pad by a random value that is contained in the ``list``) or a ``StochasticParameter`` (sample the amount to pad from that parameter). percent : None or int or float or imgaug.parameters.StochasticParameter or tuple, optional The number of pixels to pad on each side of the image given as a *fraction* of the image height/width. E.g. if this is set to ``0.1``, the augmenter will always pad ``10%`` of the image's height at both the top and the bottom (both ``10%`` each), as well as ``10%`` of the width at the right and left. Expected value range is ``[0.0, inf)``. Either this or the parameter `px` may be set, not both at the same time. * If ``None``, then fraction-based padding will not be used. * If ``number``, then that fraction will always be padded. * If ``StochasticParameter``, then that parameter will be used for each image. Four samples will be drawn per image (top, right, bottom, left). If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of two ``float`` s with values ``a`` and ``b``, then each side will be padded by a random fraction sampled uniformly per image and side from the interval ``[a, b]``. If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of four entries, then the entries represent top, right, bottom, left. Each entry may be a single ``float`` (always pad by exactly that fraction), a ``tuple`` of two ``float`` s ``a`` and ``b`` (pad by a fraction from ``[a, b]``), a ``list`` of ``float`` s (pad by a random value that is contained in the list) or a ``StochasticParameter`` (sample the percentage to pad from that parameter). pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional Padding mode to use. The available modes match the numpy padding modes, i.e. ``constant``, ``edge``, ``linear_ramp``, ``maximum``, ``median``, ``minimum``, ``reflect``, ``symmetric``, ``wrap``. The modes ``constant`` and ``linear_ramp`` use extra values, which are provided by ``pad_cval`` when necessary. See :func:`~imgaug.imgaug.pad` for more details. * If ``imgaug.ALL``, then a random mode from all available modes will be sampled per image. * If a ``str``, it will be used as the pad mode for all images. * If a ``list`` of ``str``, a random one of these will be sampled per image and used as the mode. * If ``StochasticParameter``, a random mode will be sampled from this parameter per image. pad_cval : number or tuple of number list of number or imgaug.parameters.StochasticParameter, optional The constant value to use if the pad mode is ``constant`` or the end value to use if the mode is ``linear_ramp``. See :func:`~imgaug.imgaug.pad` for more details. * If ``number``, then that value will be used. * If a ``tuple`` of two ``number`` s and at least one of them is a ``float``, then a random number will be uniformly sampled per image from the continuous interval ``[a, b]`` and used as the value. If both ``number`` s are ``int`` s, the interval is discrete. * If a ``list`` of ``number``, then a random value will be chosen from the elements of the ``list`` and used as the value. * If ``StochasticParameter``, a random value will be sampled from that parameter per image. keep_size : bool, optional After padding, the result image will usually have a different height/width compared to the original input image. If this parameter is set to ``True``, then the padded image will be resized to the input image's size, i.e. the augmenter's output shape is always identical to the input shape. sample_independently : bool, optional If ``False`` *and* the values for `px`/`percent` result in exactly *one* probability distribution for all image sides, only one single value will be sampled from that probability distribution and used for all sides. I.e. the pad amount then is the same for all sides. If ``True``, four values will be sampled independently, one per side. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.Pad(px=(0, 10)) Pad each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. The padding happens by zero-padding, i.e. it adds black pixels (default setting). >>> aug = iaa.Pad(px=(0, 10), pad_mode="edge") Pad each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. The padding uses the ``edge`` mode from numpy's pad function, i.e. the pixel colors around the image sides are repeated. >>> aug = iaa.Pad(px=(0, 10), pad_mode=["constant", "edge"]) Similar to the previous example, but uses zero-padding (``constant``) for half of the images and ``edge`` padding for the other half. >>> aug = iaa.Pad(px=(0, 10), pad_mode=ia.ALL, pad_cval=(0, 255)) Similar to the previous example, but uses any available padding mode. In case the padding mode ends up being ``constant`` or ``linear_ramp``, and random intensity is uniformly sampled (once per image) from the discrete interval ``[0..255]`` and used as the intensity of the new pixels. >>> aug = iaa.Pad(px=(0, 10), sample_independently=False) Pad each side by a random pixel value sampled uniformly once per image from the discrete interval ``[0..10]``. Each sampled value is used for *all* sides of the corresponding image. >>> aug = iaa.Pad(px=(0, 10), keep_size=False) Pad each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. Afterwards, do **not** resize the padded image back to the input image's size. This will increase the image's height and width by a maximum of ``20`` pixels. >>> aug = iaa.Pad(px=((0, 10), (0, 5), (0, 10), (0, 5))) Pad the top and bottom by a random pixel value sampled uniformly from the discrete interval ``[0..10]``. Pad the left and right analogously by a random value sampled from ``[0..5]``. Each value is always sampled independently. >>> aug = iaa.Pad(percent=(0, 0.1)) Pad each side by a random fraction sampled uniformly from the continuous interval ``[0.0, 0.10]``. The fraction is sampled once per image and side. E.g. a sampled fraction of ``0.1`` for the top side would pad by ``0.1*H``, where ``H`` is the height of the input image. >>> aug = iaa.Pad( >>> percent=([0.05, 0.1], [0.05, 0.1], [0.05, 0.1], [0.05, 0.1])) Pads each side by either ``5%`` or ``10%``. The values are sampled once per side and image. """ def __init__(self, px=None, percent=None, pad_mode="constant", pad_cval=0, keep_size=True, sample_independently=True, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): def recursive_validate(value): if value is None: return value if ia.is_single_number(value): assert value >= 0, "Expected value >0, got %.4f" % (value,) return value if isinstance(value, iap.StochasticParameter): return value if isinstance(value, tuple): return tuple([recursive_validate(v_) for v_ in value]) if isinstance(value, list): return [recursive_validate(v_) for v_ in value] raise Exception( "Expected None or int or float or StochasticParameter or " "list or tuple, got %s." % (type(value),)) if px is None and percent is None: percent = (0.0, 0.1) px = recursive_validate(px) percent = recursive_validate(percent) super(Pad, self).__init__( px=px, percent=percent, pad_mode=pad_mode, pad_cval=pad_cval, keep_size=keep_size, sample_independently=sample_independently, seed=seed, name=name, random_state=random_state, deterministic=deterministic) class Crop(CropAndPad): """Crop images, i.e. remove columns/rows of pixels at the sides of images. This augmenter allows to extract smaller-sized subimages from given full-sized input images. The number of pixels to cut off may be defined in absolute values or as fractions of the image sizes. This augmenter will never crop images below a height or width of ``1``. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropAndPad`. Parameters ---------- px : None or int or imgaug.parameters.StochasticParameter or tuple, optional The number of pixels to crop on each side of the image. Expected value range is ``[0, inf)``. Either this or the parameter `percent` may be set, not both at the same time. * If ``None``, then pixel-based cropping will not be used. * If ``int``, then that exact number of pixels will always be cropped. * If ``StochasticParameter``, then that parameter will be used for each image. Four samples will be drawn per image (top, right, bottom, left), unless `sample_independently` is set to ``False``, as then only one value will be sampled per image and used for all sides. * If a ``tuple`` of two ``int`` s with values ``a`` and ``b``, then each side will be cropped by a random amount sampled uniformly per image and side from the inteval ``[a, b]``. If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of four entries, then the entries represent top, right, bottom, left. Each entry may be a single ``int`` (always crop by exactly that value), a ``tuple`` of two ``int`` s ``a`` and ``b`` (crop by an amount within ``[a, b]``), a ``list`` of ``int`` s (crop by a random value that is contained in the ``list``) or a ``StochasticParameter`` (sample the amount to crop from that parameter). percent : None or int or float or imgaug.parameters.StochasticParameter or tuple, optional The number of pixels to crop on each side of the image given as a *fraction* of the image height/width. E.g. if this is set to ``0.1``, the augmenter will always crop ``10%`` of the image's height at both the top and the bottom (both ``10%`` each), as well as ``10%`` of the width at the right and left. Expected value range is ``[0.0, 1.0)``. Either this or the parameter `px` may be set, not both at the same time. * If ``None``, then fraction-based cropping will not be used. * If ``number``, then that fraction will always be cropped. * If ``StochasticParameter``, then that parameter will be used for each image. Four samples will be drawn per image (top, right, bottom, left). If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of two ``float`` s with values ``a`` and ``b``, then each side will be cropped by a random fraction sampled uniformly per image and side from the interval ``[a, b]``. If however `sample_independently` is set to ``False``, only one value will be sampled per image and used for all sides. * If a ``tuple`` of four entries, then the entries represent top, right, bottom, left. Each entry may be a single ``float`` (always crop by exactly that fraction), a ``tuple`` of two ``float`` s ``a`` and ``b`` (crop by a fraction from ``[a, b]``), a ``list`` of ``float`` s (crop by a random value that is contained in the list) or a ``StochasticParameter`` (sample the percentage to crop from that parameter). keep_size : bool, optional After cropping, the result image will usually have a different height/width compared to the original input image. If this parameter is set to ``True``, then the cropped image will be resized to the input image's size, i.e. the augmenter's output shape is always identical to the input shape. sample_independently : bool, optional If ``False`` *and* the values for `px`/`percent` result in exactly *one* probability distribution for all image sides, only one single value will be sampled from that probability distribution and used for all sides. I.e. the crop amount then is the same for all sides. If ``True``, four values will be sampled independently, one per side. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.Crop(px=(0, 10)) Crop each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. >>> aug = iaa.Crop(px=(0, 10), sample_independently=False) Crop each side by a random pixel value sampled uniformly once per image from the discrete interval ``[0..10]``. Each sampled value is used for *all* sides of the corresponding image. >>> aug = iaa.Crop(px=(0, 10), keep_size=False) Crop each side by a random pixel value sampled uniformly per image and side from the discrete interval ``[0..10]``. Afterwards, do **not** resize the cropped image back to the input image's size. This will decrease the image's height and width by a maximum of ``20`` pixels. >>> aug = iaa.Crop(px=((0, 10), (0, 5), (0, 10), (0, 5))) Crop the top and bottom by a random pixel value sampled uniformly from the discrete interval ``[0..10]``. Crop the left and right analogously by a random value sampled from ``[0..5]``. Each value is always sampled independently. >>> aug = iaa.Crop(percent=(0, 0.1)) Crop each side by a random fraction sampled uniformly from the continuous interval ``[0.0, 0.10]``. The fraction is sampled once per image and side. E.g. a sampled fraction of ``0.1`` for the top side would crop by ``0.1*H``, where ``H`` is the height of the input image. >>> aug = iaa.Crop( >>> percent=([0.05, 0.1], [0.05, 0.1], [0.05, 0.1], [0.05, 0.1])) Crops each side by either ``5%`` or ``10%``. The values are sampled once per side and image. """ def __init__(self, px=None, percent=None, keep_size=True, sample_independently=True, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): def recursive_negate(value): if value is None: return value if ia.is_single_number(value): assert value >= 0, "Expected value >0, got %.4f." % (value,) return -value if isinstance(value, iap.StochasticParameter): return iap.Multiply(value, -1) if isinstance(value, tuple): return tuple([recursive_negate(v_) for v_ in value]) if isinstance(value, list): return [recursive_negate(v_) for v_ in value] raise Exception( "Expected None or int or float or StochasticParameter or " "list or tuple, got %s." % (type(value),)) if px is None and percent is None: percent = (0.0, 0.1) px = recursive_negate(px) percent = recursive_negate(percent) super(Crop, self).__init__( px=px, percent=percent, keep_size=keep_size, sample_independently=sample_independently, seed=seed, name=name, random_state=random_state, deterministic=deterministic) # TODO maybe rename this to PadToMinimumSize? # TODO this is very similar to CropAndPad, maybe add a way to generate crop # values imagewise via a callback in in CropAndPad? # TODO why is padding mode and cval here called pad_mode, pad_cval but in other # cases mode/cval? class PadToFixedSize(meta.Augmenter): """Pad images to a predefined minimum width and/or height. If images are already at the minimum width/height or are larger, they will not be padded. Note that this also means that images will not be cropped if they exceed the required width/height. The augmenter randomly decides per image how to distribute the required padding amounts over the image axis. E.g. if 2px have to be padded on the left or right to reach the required width, the augmenter will sometimes add 2px to the left and 0px to the right, sometimes add 2px to the right and 0px to the left and sometimes add 1px to both sides. Set `position` to ``center`` to prevent that. **Supported dtypes**: See :func:`~imgaug.augmenters.size.pad`. Parameters ---------- width : int or None Pad images up to this minimum width. If ``None``, image widths will not be altered. height : int or None Pad images up to this minimum height. If ``None``, image heights will not be altered. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.CropAndPad.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.CropAndPad.__init__`. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional Sets the center point of the padding, which determines how the required padding amounts are distributed to each side. For a ``tuple`` ``(a, b)``, both ``a`` and ``b`` are expected to be in range ``[0.0, 1.0]`` and describe the fraction of padding applied to the left/right (low/high values for ``a``) and the fraction of padding applied to the top/bottom (low/high values for ``b``). A padding position at ``(0.5, 0.5)`` would be the center of the image and distribute the padding equally to all sides. A padding position at ``(0.0, 1.0)`` would be the left-bottom and would apply 100% of the required padding to the bottom and left sides of the image so that the bottom left corner becomes more and more the new image center (depending on how much is padded). * If string ``uniform`` then the share of padding is randomly and uniformly distributed over each side. Equivalent to ``(Uniform(0.0, 1.0), Uniform(0.0, 1.0))``. * If string ``normal`` then the share of padding is distributed based on a normal distribution, leading to a focus on the center of the images. Equivalent to ``(Clip(Normal(0.5, 0.45/2), 0, 1), Clip(Normal(0.5, 0.45/2), 0, 1))``. * If string ``center`` then center point of the padding is identical to the image center. Equivalent to ``(0.5, 0.5)``. * If a string matching regex ``^(left|center|right)-(top|center|bottom)$``, e.g. ``left-top`` or ``center-bottom`` then sets the center point of the padding to the X-Y position matching that description. * If a tuple of float, then expected to have exactly two entries between ``0.0`` and ``1.0``, which will always be used as the combination the position matching (x, y) form. * If a ``StochasticParameter``, then that parameter will be queried once per call to ``augment_*()`` to get ``Nx2`` center positions in ``(x, y)`` form (with ``N`` the number of images). * If a ``tuple`` of ``StochasticParameter``, then expected to have exactly two entries that will both be queried per call to ``augment_*()``, each for ``(N,)`` values, to get the center positions. First parameter is used for ``x`` coordinates, second for ``y`` coordinates. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.PadToFixedSize(width=100, height=100) For image sides smaller than ``100`` pixels, pad to ``100`` pixels. Do nothing for the other edges. The padding is randomly (uniformly) distributed over the sides, so that e.g. sometimes most of the required padding is applied to the left, sometimes to the right (analogous top/bottom). >>> aug = iaa.PadToFixedSize(width=100, height=100, position="center") For image sides smaller than ``100`` pixels, pad to ``100`` pixels. Do nothing for the other image sides. The padding is always equally distributed over the left/right and top/bottom sides. >>> aug = iaa.PadToFixedSize(width=100, height=100, pad_mode=ia.ALL) For image sides smaller than ``100`` pixels, pad to ``100`` pixels and use any possible padding mode for that. Do nothing for the other image sides. The padding is always equally distributed over the left/right and top/bottom sides. >>> aug = iaa.Sequential([ >>> iaa.PadToFixedSize(width=100, height=100), >>> iaa.CropToFixedSize(width=100, height=100) >>> ]) Pad images smaller than ``100x100`` until they reach ``100x100``. Analogously, crop images larger than ``100x100`` until they reach ``100x100``. The output images therefore have a fixed size of ``100x100``. """ def __init__(self, width, height, pad_mode="constant", pad_cval=0, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(PadToFixedSize, self).__init__( seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.size = (width, height) # Position of where to pad. The further to the top left this is, the # larger the share of pixels that will be added to the top and left # sides. I.e. set to (Deterministic(0.0), Deterministic(0.0)) to only # add at the top and left, (Deterministic(1.0), Deterministic(1.0)) # to only add at the bottom right. Analogously (0.5, 0.5) pads equally # on both axis, (0.0, 1.0) pads left and bottom, (1.0, 0.0) pads right # and top. self.position = _handle_position_parameter(position) self.pad_mode = _handle_pad_mode_param(pad_mode) # TODO enable ALL here like in eg Affine self.pad_cval = iap.handle_discrete_param( pad_cval, "pad_cval", value_range=None, tuple_to_uniform=True, list_to_choice=True, allow_floats=True) # set these to None to use the same values as sampled for the # images (not tested) self._pad_mode_heatmaps = "constant" self._pad_mode_segmentation_maps = "constant" self._pad_cval_heatmaps = 0.0 self._pad_cval_segmentation_maps = 0 # Added in 0.4.0. def _augment_batch_(self, batch, random_state, parents, hooks): # Providing the whole batch to _draw_samples() would not be necessary # for this augmenter. The number of rows would be sufficient. This # formulation however enables derived augmenters to use rowwise shapes # without having to compute them here for this augmenter. samples = self._draw_samples(batch, random_state) if batch.images is not None: batch.images = self._augment_images_by_samples(batch.images, samples) if batch.heatmaps is not None: batch.heatmaps = self._augment_maps_by_samples( batch.heatmaps, samples, self._pad_mode_heatmaps, self._pad_cval_heatmaps) if batch.segmentation_maps is not None: batch.segmentation_maps = self._augment_maps_by_samples( batch.segmentation_maps, samples, self._pad_mode_heatmaps, self._pad_cval_heatmaps) for augm_name in ["keypoints", "bounding_boxes", "polygons", "line_strings"]: augm_value = getattr(batch, augm_name) if augm_value is not None: func = functools.partial( self._augment_keypoints_by_samples, samples=samples) cbaois = self._apply_to_cbaois_as_keypoints(augm_value, func) setattr(batch, augm_name, cbaois) return batch # Added in 0.4.0. def _augment_images_by_samples(self, images, samples): result = [] sizes, pad_xs, pad_ys, pad_modes, pad_cvals = samples for i, (image, size) in enumerate(zip(images, sizes)): width_min, height_min = size height_image, width_image = image.shape[:2] paddings = self._calculate_paddings(height_image, width_image, height_min, width_min, pad_xs[i], pad_ys[i]) image = _crop_and_pad_arr( image, (0, 0, 0, 0), paddings, pad_modes[i], pad_cvals[i], keep_size=False) result.append(image) # TODO result is always a list. Should this be converted to an array # if possible (not guaranteed that all images have same size, # some might have been larger than desired height/width) return result # Added in 0.4.0. def _augment_keypoints_by_samples(self, keypoints_on_images, samples): result = [] sizes, pad_xs, pad_ys, _, _ = samples for i, (kpsoi, size) in enumerate(zip(keypoints_on_images, sizes)): width_min, height_min = size height_image, width_image = kpsoi.shape[:2] paddings_img = self._calculate_paddings(height_image, width_image, height_min, width_min, pad_xs[i], pad_ys[i]) keypoints_padded = _crop_and_pad_kpsoi_( kpsoi, (0, 0, 0, 0), paddings_img, keep_size=False) result.append(keypoints_padded) return result # Added in 0.4.0. def _augment_maps_by_samples(self, augmentables, samples, pad_mode, pad_cval): sizes, pad_xs, pad_ys, pad_modes, pad_cvals = samples for i, (augmentable, size) in enumerate(zip(augmentables, sizes)): width_min, height_min = size height_img, width_img = augmentable.shape[:2] paddings_img = self._calculate_paddings( height_img, width_img, height_min, width_min, pad_xs[i], pad_ys[i]) # TODO for the previous method (and likely the new/current one # too): # for 30x30 padded to 32x32 with 15x15 heatmaps this results # in paddings of 1 on each side (assuming # position=(0.5, 0.5)) giving 17x17 heatmaps when they should # be 16x16. Error is due to each side getting projected 0.5 # padding which is rounded to 1. This doesn't seem right. augmentables[i] = _crop_and_pad_hms_or_segmaps_( augmentables[i], (0, 0, 0, 0), paddings_img, pad_mode=pad_mode if pad_mode is not None else pad_modes[i], pad_cval=pad_cval if pad_cval is not None else pad_cvals[i], keep_size=False) return augmentables def _draw_samples(self, batch, random_state): nb_images = batch.nb_rows rngs = random_state.duplicate(4) if isinstance(self.position, tuple): pad_xs = self.position[0].draw_samples(nb_images, random_state=rngs[0]) pad_ys = self.position[1].draw_samples(nb_images, random_state=rngs[1]) else: pads = self.position.draw_samples((nb_images, 2), random_state=rngs[0]) pad_xs = pads[:, 0] pad_ys = pads[:, 1] pad_modes = self.pad_mode.draw_samples(nb_images, random_state=rngs[2]) pad_cvals = self.pad_cval.draw_samples(nb_images, random_state=rngs[3]) # We return here the sizes even though they are static as it allows # derived augmenters to define image-specific heights/widths. return [self.size] * nb_images, pad_xs, pad_ys, pad_modes, pad_cvals @classmethod def _calculate_paddings(cls, height_image, width_image, height_min, width_min, pad_xs_i, pad_ys_i): pad_top = 0 pad_right = 0 pad_bottom = 0 pad_left = 0 if width_min is not None and width_image < width_min: pad_total_x = width_min - width_image pad_left = int((1-pad_xs_i) * pad_total_x) pad_right = pad_total_x - pad_left if height_min is not None and height_image < height_min: pad_total_y = height_min - height_image pad_top = int((1-pad_ys_i) * pad_total_y) pad_bottom = pad_total_y - pad_top return pad_top, pad_right, pad_bottom, pad_left def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.size[0], self.size[1], self.pad_mode, self.pad_cval, self.position] class CenterPadToFixedSize(PadToFixedSize): """Pad images equally on all sides up to given minimum heights/widths. This is an alias for :class:`~imgaug.augmenters.size.PadToFixedSize` with ``position="center"``. It spreads the pad amounts equally over all image sides, while :class:`~imgaug.augmenters.size.PadToFixedSize` by defaults spreads them randomly. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- width : int or None See :func:`PadToFixedSize.__init__`. height : int or None See :func:`PadToFixedSize.__init__`. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`PadToFixedSize.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`PadToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CenterPadToFixedSize(height=20, width=30) Create an augmenter that pads images up to ``20x30``, with the padded rows added *equally* on the top and bottom (analogous for the padded columns). """ # Added in 0.4.0. def __init__(self, width, height, pad_mode="constant", pad_cval=0, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterPadToFixedSize, self).__init__( width=width, height=height, pad_mode=pad_mode, pad_cval=pad_cval, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) # TODO maybe rename this to CropToMaximumSize ? # TODO this is very similar to CropAndPad, maybe add a way to generate crop # values imagewise via a callback in in CropAndPad? # TODO add crop() function in imgaug, similar to pad class CropToFixedSize(meta.Augmenter): """Crop images down to a predefined maximum width and/or height. If images are already at the maximum width/height or are smaller, they will not be cropped. Note that this also means that images will not be padded if they are below the required width/height. The augmenter randomly decides per image how to distribute the required cropping amounts over the image axis. E.g. if 2px have to be cropped on the left or right to reach the required width, the augmenter will sometimes remove 2px from the left and 0px from the right, sometimes remove 2px from the right and 0px from the left and sometimes remove 1px from both sides. Set `position` to ``center`` to prevent that. **Supported dtypes**: * ``uint8``: yes; fully tested * ``uint16``: yes; tested * ``uint32``: yes; tested * ``uint64``: yes; tested * ``int8``: yes; tested * ``int16``: yes; tested * ``int32``: yes; tested * ``int64``: yes; tested * ``float16``: yes; tested * ``float32``: yes; tested * ``float64``: yes; tested * ``float128``: yes; tested * ``bool``: yes; tested Parameters ---------- width : int or None Crop images down to this maximum width. If ``None``, image widths will not be altered. height : int or None Crop images down to this maximum height. If ``None``, image heights will not be altered. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional Sets the center point of the cropping, which determines how the required cropping amounts are distributed to each side. For a ``tuple`` ``(a, b)``, both ``a`` and ``b`` are expected to be in range ``[0.0, 1.0]`` and describe the fraction of cropping applied to the left/right (low/high values for ``a``) and the fraction of cropping applied to the top/bottom (low/high values for ``b``). A cropping position at ``(0.5, 0.5)`` would be the center of the image and distribute the cropping equally over all sides. A cropping position at ``(1.0, 0.0)`` would be the right-top and would apply 100% of the required cropping to the right and top sides of the image. * If string ``uniform`` then the share of cropping is randomly and uniformly distributed over each side. Equivalent to ``(Uniform(0.0, 1.0), Uniform(0.0, 1.0))``. * If string ``normal`` then the share of cropping is distributed based on a normal distribution, leading to a focus on the center of the images. Equivalent to ``(Clip(Normal(0.5, 0.45/2), 0, 1), Clip(Normal(0.5, 0.45/2), 0, 1))``. * If string ``center`` then center point of the cropping is identical to the image center. Equivalent to ``(0.5, 0.5)``. * If a string matching regex ``^(left|center|right)-(top|center|bottom)$``, e.g. ``left-top`` or ``center-bottom`` then sets the center point of the cropping to the X-Y position matching that description. * If a tuple of float, then expected to have exactly two entries between ``0.0`` and ``1.0``, which will always be used as the combination the position matching (x, y) form. * If a ``StochasticParameter``, then that parameter will be queried once per call to ``augment_*()`` to get ``Nx2`` center positions in ``(x, y)`` form (with ``N`` the number of images). * If a ``tuple`` of ``StochasticParameter``, then expected to have exactly two entries that will both be queried per call to ``augment_*()``, each for ``(N,)`` values, to get the center positions. First parameter is used for ``x`` coordinates, second for ``y`` coordinates. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CropToFixedSize(width=100, height=100) For image sides larger than ``100`` pixels, crop to ``100`` pixels. Do nothing for the other sides. The cropping amounts are randomly (and uniformly) distributed over the sides of the image. >>> aug = iaa.CropToFixedSize(width=100, height=100, position="center") For sides larger than ``100`` pixels, crop to ``100`` pixels. Do nothing for the other sides. The cropping amounts are always equally distributed over the left/right sides of the image (and analogously for top/bottom). >>> aug = iaa.Sequential([ >>> iaa.PadToFixedSize(width=100, height=100), >>> iaa.CropToFixedSize(width=100, height=100) >>> ]) Pad images smaller than ``100x100`` until they reach ``100x100``. Analogously, crop images larger than ``100x100`` until they reach ``100x100``. The output images therefore have a fixed size of ``100x100``. """ def __init__(self, width, height, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CropToFixedSize, self).__init__( seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.size = (width, height) # Position of where to crop. The further to the top left this is, # the larger the share of pixels that will be cropped from the top # and left sides. I.e. set to (Deterministic(0.0), Deterministic(0.0)) # to only crop at the top and left, # (Deterministic(1.0), Deterministic(1.0)) to only crop at the bottom # right. Analogously (0.5, 0.5) crops equally on both axis, # (0.0, 1.0) crops left and bottom, (1.0, 0.0) crops right and top. self.position = _handle_position_parameter(position) # Added in 0.4.0. def _augment_batch_(self, batch, random_state, parents, hooks): # Providing the whole batch to _draw_samples() would not be necessary # for this augmenter. The number of rows would be sufficient. This # formulation however enables derived augmenters to use rowwise shapes # without having to compute them here for this augmenter. samples = self._draw_samples(batch, random_state) if batch.images is not None: batch.images = self._augment_images_by_samples(batch.images, samples) if batch.heatmaps is not None: batch.heatmaps = self._augment_maps_by_samples( batch.heatmaps, samples) if batch.segmentation_maps is not None: batch.segmentation_maps = self._augment_maps_by_samples( batch.segmentation_maps, samples) for augm_name in ["keypoints", "bounding_boxes", "polygons", "line_strings"]: augm_value = getattr(batch, augm_name) if augm_value is not None: func = functools.partial( self._augment_keypoints_by_samples, samples=samples) cbaois = self._apply_to_cbaois_as_keypoints(augm_value, func) setattr(batch, augm_name, cbaois) return batch # Added in 0.4.0. def _augment_images_by_samples(self, images, samples): result = [] sizes, offset_xs, offset_ys = samples for i, (image, size) in enumerate(zip(images, sizes)): w, h = size height_image, width_image = image.shape[0:2] croppings = self._calculate_crop_amounts( height_image, width_image, h, w, offset_ys[i], offset_xs[i]) image_cropped = _crop_and_pad_arr(image, croppings, (0, 0, 0, 0), keep_size=False) result.append(image_cropped) return result # Added in 0.4.0. def _augment_keypoints_by_samples(self, kpsois, samples): result = [] sizes, offset_xs, offset_ys = samples for i, (kpsoi, size) in enumerate(zip(kpsois, sizes)): w, h = size height_image, width_image = kpsoi.shape[0:2] croppings_img = self._calculate_crop_amounts( height_image, width_image, h, w, offset_ys[i], offset_xs[i]) kpsoi_cropped = _crop_and_pad_kpsoi_( kpsoi, croppings_img, (0, 0, 0, 0), keep_size=False) result.append(kpsoi_cropped) return result # Added in 0.4.0. def _augment_maps_by_samples(self, augmentables, samples): sizes, offset_xs, offset_ys = samples for i, (augmentable, size) in enumerate(zip(augmentables, sizes)): w, h = size height_image, width_image = augmentable.shape[0:2] croppings_img = self._calculate_crop_amounts( height_image, width_image, h, w, offset_ys[i], offset_xs[i]) augmentables[i] = _crop_and_pad_hms_or_segmaps_( augmentable, croppings_img, (0, 0, 0, 0), keep_size=False) return augmentables @classmethod def _calculate_crop_amounts(cls, height_image, width_image, height_max, width_max, offset_y, offset_x): crop_top = 0 crop_right = 0 crop_bottom = 0 crop_left = 0 if height_max is not None and height_image > height_max: crop_top = int(offset_y * (height_image - height_max)) crop_bottom = height_image - height_max - crop_top if width_max is not None and width_image > width_max: crop_left = int(offset_x * (width_image - width_max)) crop_right = width_image - width_max - crop_left return crop_top, crop_right, crop_bottom, crop_left def _draw_samples(self, batch, random_state): nb_images = batch.nb_rows rngs = random_state.duplicate(2) if isinstance(self.position, tuple): offset_xs = self.position[0].draw_samples(nb_images, random_state=rngs[0]) offset_ys = self.position[1].draw_samples(nb_images, random_state=rngs[1]) else: offsets = self.position.draw_samples((nb_images, 2), random_state=rngs[0]) offset_xs = offsets[:, 0] offset_ys = offsets[:, 1] offset_xs = 1.0 - offset_xs offset_ys = 1.0 - offset_ys # We return here the sizes even though they are static as it allows # derived augmenters to define image-specific heights/widths. return [self.size] * nb_images, offset_xs, offset_ys def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.size[0], self.size[1], self.position] class CenterCropToFixedSize(CropToFixedSize): """Take a crop from the center of each image. This is an alias for :class:`~imgaug.augmenters.size.CropToFixedSize` with ``position="center"``. .. note:: If images already have a width and/or height below the provided width and/or height then this augmenter will do nothing for the respective axis. Hence, resulting images can be smaller than the provided axis sizes. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- width : int or None See :func:`CropToFixedSize.__init__`. height : int or None See :func:`CropToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> crop = iaa.CenterCropToFixedSize(height=20, width=10) Create an augmenter that takes ``20x10`` sized crops from the center of images. """ # Added in 0.4.0. def __init__(self, width, height, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterCropToFixedSize, self).__init__( width=width, height=height, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class CropToMultiplesOf(CropToFixedSize): """Crop images down until their height/width is a multiple of a value. .. note:: For a given axis size ``A`` and multiple ``M``, if ``A`` is in the interval ``[0 .. M]``, the axis will not be changed. As a result, this augmenter can still produce axis sizes that are not multiples of the given values. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- width_multiple : int or None Multiple for the width. Images will be cropped down until their width is a multiple of this value. If ``None``, image widths will not be altered. height_multiple : int or None Multiple for the height. Images will be cropped down until their height is a multiple of this value. If ``None``, image heights will not be altered. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`CropToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CropToMultiplesOf(height_multiple=10, width_multiple=6) Create an augmenter that crops images to multiples of ``10`` along the y-axis (i.e. 10, 20, 30, ...) and to multiples of ``6`` along the x-axis (i.e. 6, 12, 18, ...). The rows to be cropped will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_multiple, height_multiple, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CropToMultiplesOf, self).__init__( width=None, height=None, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.width_multiple = width_multiple self.height_multiple = height_multiple # Added in 0.4.0. def _draw_samples(self, batch, random_state): _sizes, offset_xs, offset_ys = super( CropToMultiplesOf, self )._draw_samples(batch, random_state) shapes = batch.get_rowwise_shapes() sizes = [] for shape in shapes: height, width = shape[0:2] croppings = compute_croppings_to_reach_multiples_of( shape, height_multiple=self.height_multiple, width_multiple=self.width_multiple) # TODO change that # note that these are not in the same order as shape tuples # in CropToFixedSize new_size = ( width - croppings[1] - croppings[3], height - croppings[0] - croppings[2] ) sizes.append(new_size) return sizes, offset_xs, offset_ys # Added in 0.4.0. def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.width_multiple, self.height_multiple, self.position] class CenterCropToMultiplesOf(CropToMultiplesOf): """Crop images equally on all sides until H/W are multiples of given values. This is the same as :class:`~imgaug.augmenters.size.CropToMultiplesOf`, but uses ``position="center"`` by default, which spreads the crop amounts equally over all image sides, while :class:`~imgaug.augmenters.size.CropToMultiplesOf` by default spreads them randomly. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- width_multiple : int or None See :func:`CropToMultiplesOf.__init__`. height_multiple : int or None See :func:`CropToMultiplesOf.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CenterCropToMultiplesOf(height_multiple=10, width_multiple=6) Create an augmenter that crops images to multiples of ``10`` along the y-axis (i.e. 10, 20, 30, ...) and to multiples of ``6`` along the x-axis (i.e. 6, 12, 18, ...). The rows to be cropped will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_multiple, height_multiple, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterCropToMultiplesOf, self).__init__( width_multiple=width_multiple, height_multiple=height_multiple, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class PadToMultiplesOf(PadToFixedSize): """Pad images until their height/width is a multiple of a value. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- width_multiple : int or None Multiple for the width. Images will be padded until their width is a multiple of this value. If ``None``, image widths will not be altered. height_multiple : int or None Multiple for the height. Images will be padded until their height is a multiple of this value. If ``None``, image heights will not be altered. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`PadToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.PadToMultiplesOf(height_multiple=10, width_multiple=6) Create an augmenter that pads images to multiples of ``10`` along the y-axis (i.e. 10, 20, 30, ...) and to multiples of ``6`` along the x-axis (i.e. 6, 12, 18, ...). The rows to be padded will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_multiple, height_multiple, pad_mode="constant", pad_cval=0, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(PadToMultiplesOf, self).__init__( width=None, height=None, pad_mode=pad_mode, pad_cval=pad_cval, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.width_multiple = width_multiple self.height_multiple = height_multiple # Added in 0.4.0. def _draw_samples(self, batch, random_state): _sizes, pad_xs, pad_ys, pad_modes, pad_cvals = super( PadToMultiplesOf, self )._draw_samples(batch, random_state) shapes = batch.get_rowwise_shapes() sizes = [] for shape in shapes: height, width = shape[0:2] paddings = compute_paddings_to_reach_multiples_of( shape, height_multiple=self.height_multiple, width_multiple=self.width_multiple) # TODO change that # note that these are not in the same order as shape tuples # in PadToFixedSize new_size = ( width + paddings[1] + paddings[3], height + paddings[0] + paddings[2] ) sizes.append(new_size) return sizes, pad_xs, pad_ys, pad_modes, pad_cvals # Added in 0.4.0. def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.width_multiple, self.height_multiple, self.pad_mode, self.pad_cval, self.position] class CenterPadToMultiplesOf(PadToMultiplesOf): """Pad images equally on all sides until H/W are multiples of given values. This is the same as :class:`~imgaug.augmenters.size.PadToMultiplesOf`, but uses ``position="center"`` by default, which spreads the pad amounts equally over all image sides, while :class:`~imgaug.augmenters.size.PadToMultiplesOf` by default spreads them randomly. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- width_multiple : int or None See :func:`PadToMultiplesOf.__init__`. height_multiple : int or None See :func:`PadToMultiplesOf.__init__`. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToMultiplesOf.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToMultiplesOf.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CenterPadToMultiplesOf(height_multiple=10, width_multiple=6) Create an augmenter that pads images to multiples of ``10`` along the y-axis (i.e. 10, 20, 30, ...) and to multiples of ``6`` along the x-axis (i.e. 6, 12, 18, ...). The rows to be padded will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_multiple, height_multiple, pad_mode="constant", pad_cval=0, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterPadToMultiplesOf, self).__init__( width_multiple=width_multiple, height_multiple=height_multiple, pad_mode=pad_mode, pad_cval=pad_cval, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class CropToPowersOf(CropToFixedSize): """Crop images until their height/width is a power of a base. This augmenter removes pixels from an axis with size ``S`` leading to the new size ``S'`` until ``S' = B^E`` is fulfilled, where ``B`` is a provided base (e.g. ``2``) and ``E`` is an exponent from the discrete interval ``[1 .. inf)``. .. note:: This augmenter does nothing for axes with size less than ``B^1 = B``. If you have images with ``S < B^1``, it is recommended to combine this augmenter with a padding augmenter that pads each axis up to ``B``. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- width_base : int or None Base for the width. Images will be cropped down until their width fulfills ``width' = width_base ^ E`` with ``E`` being any natural number. If ``None``, image widths will not be altered. height_base : int or None Base for the height. Images will be cropped down until their height fulfills ``height' = height_base ^ E`` with ``E`` being any natural number. If ``None``, image heights will not be altered. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`CropToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CropToPowersOf(height_base=3, width_base=2) Create an augmenter that crops each image down to powers of ``3`` along the y-axis (i.e. 3, 9, 27, ...) and powers of ``2`` along the x-axis (i.e. 2, 4, 8, 16, ...). The rows to be cropped will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_base, height_base, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CropToPowersOf, self).__init__( width=None, height=None, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.width_base = width_base self.height_base = height_base # Added in 0.4.0. def _draw_samples(self, batch, random_state): _sizes, offset_xs, offset_ys = super( CropToPowersOf, self )._draw_samples(batch, random_state) shapes = batch.get_rowwise_shapes() sizes = [] for shape in shapes: height, width = shape[0:2] croppings = compute_croppings_to_reach_powers_of( shape, height_base=self.height_base, width_base=self.width_base) # TODO change that # note that these are not in the same order as shape tuples # in CropToFixedSize new_size = ( width - croppings[1] - croppings[3], height - croppings[0] - croppings[2] ) sizes.append(new_size) return sizes, offset_xs, offset_ys # Added in 0.4.0. def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.width_base, self.height_base, self.position] class CenterCropToPowersOf(CropToPowersOf): """Crop images equally on all sides until H/W is a power of a base. This is the same as :class:`~imgaug.augmenters.size.CropToPowersOf`, but uses ``position="center"`` by default, which spreads the crop amounts equally over all image sides, while :class:`~imgaug.augmenters.size.CropToPowersOf` by default spreads them randomly. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- width_base : int or None See :func:`CropToPowersOf.__init__`. height_base : int or None See :func:`CropToPowersOf.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CropToPowersOf(height_base=3, width_base=2) Create an augmenter that crops each image down to powers of ``3`` along the y-axis (i.e. 3, 9, 27, ...) and powers of ``2`` along the x-axis (i.e. 2, 4, 8, 16, ...). The rows to be cropped will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_base, height_base, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterCropToPowersOf, self).__init__( width_base=width_base, height_base=height_base, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class PadToPowersOf(PadToFixedSize): """Pad images until their height/width is a power of a base. This augmenter adds pixels to an axis with size ``S`` leading to the new size ``S'`` until ``S' = B^E`` is fulfilled, where ``B`` is a provided base (e.g. ``2``) and ``E`` is an exponent from the discrete interval ``[1 .. inf)``. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- width_base : int or None Base for the width. Images will be padded down until their width fulfills ``width' = width_base ^ E`` with ``E`` being any natural number. If ``None``, image widths will not be altered. height_base : int or None Base for the height. Images will be padded until their height fulfills ``height' = height_base ^ E`` with ``E`` being any natural number. If ``None``, image heights will not be altered. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`PadToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.PadToPowersOf(height_base=3, width_base=2) Create an augmenter that pads each image to powers of ``3`` along the y-axis (i.e. 3, 9, 27, ...) and powers of ``2`` along the x-axis (i.e. 2, 4, 8, 16, ...). The rows to be padded will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_base, height_base, pad_mode="constant", pad_cval=0, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(PadToPowersOf, self).__init__( width=None, height=None, pad_mode=pad_mode, pad_cval=pad_cval, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.width_base = width_base self.height_base = height_base # Added in 0.4.0. def _draw_samples(self, batch, random_state): _sizes, pad_xs, pad_ys, pad_modes, pad_cvals = super( PadToPowersOf, self )._draw_samples(batch, random_state) shapes = batch.get_rowwise_shapes() sizes = [] for shape in shapes: height, width = shape[0:2] paddings = compute_paddings_to_reach_powers_of( shape, height_base=self.height_base, width_base=self.width_base) # TODO change that # note that these are not in the same order as shape tuples # in PadToFixedSize new_size = ( width + paddings[1] + paddings[3], height + paddings[0] + paddings[2] ) sizes.append(new_size) return sizes, pad_xs, pad_ys, pad_modes, pad_cvals # Added in 0.4.0. def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.width_base, self.height_base, self.pad_mode, self.pad_cval, self.position] class CenterPadToPowersOf(PadToPowersOf): """Pad images equally on all sides until H/W is a power of a base. This is the same as :class:`~imgaug.augmenters.size.PadToPowersOf`, but uses ``position="center"`` by default, which spreads the pad amounts equally over all image sides, while :class:`~imgaug.augmenters.size.PadToPowersOf` by default spreads them randomly. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- width_base : int or None See :func:`PadToPowersOf.__init__`. height_base : int or None See :func:`PadToPowersOf.__init__`. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToPowersOf.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToPowersOf.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CenterPadToPowersOf(height_base=5, width_base=2) Create an augmenter that pads each image to powers of ``3`` along the y-axis (i.e. 3, 9, 27, ...) and powers of ``2`` along the x-axis (i.e. 2, 4, 8, 16, ...). The rows to be padded will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, width_base, height_base, pad_mode="constant", pad_cval=0, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterPadToPowersOf, self).__init__( width_base=width_base, height_base=height_base, pad_mode=pad_mode, pad_cval=pad_cval, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class CropToAspectRatio(CropToFixedSize): """Crop images until their width/height matches an aspect ratio. This augmenter removes either rows or columns until the image reaches the desired aspect ratio given in ``width / height``. The cropping operation is stopped once the desired aspect ratio is reached or the image side to crop reaches a size of ``1``. If any side of the image starts with a size of ``0``, the image will not be changed. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- aspect_ratio : number The desired aspect ratio, given as ``width/height``. E.g. a ratio of ``2.0`` denotes an image that is twice as wide as it is high. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`CropToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CropToAspectRatio(2.0) Create an augmenter that crops each image until its aspect ratio is as close as possible to ``2.0`` (i.e. two times as many pixels along the x-axis than the y-axis). The rows to be cropped will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, aspect_ratio, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CropToAspectRatio, self).__init__( width=None, height=None, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.aspect_ratio = aspect_ratio # Added in 0.4.0. def _draw_samples(self, batch, random_state): _sizes, offset_xs, offset_ys = super( CropToAspectRatio, self )._draw_samples(batch, random_state) shapes = batch.get_rowwise_shapes() sizes = [] for shape in shapes: height, width = shape[0:2] if height == 0 or width == 0: croppings = (0, 0, 0, 0) else: croppings = compute_croppings_to_reach_aspect_ratio( shape, aspect_ratio=self.aspect_ratio) # TODO change that # note that these are not in the same order as shape tuples # in CropToFixedSize new_size = ( width - croppings[1] - croppings[3], height - croppings[0] - croppings[2] ) sizes.append(new_size) return sizes, offset_xs, offset_ys # Added in 0.4.0. def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.aspect_ratio, self.position] class CenterCropToAspectRatio(CropToAspectRatio): """Crop images equally on all sides until they reach an aspect ratio. This is the same as :class:`~imgaug.augmenters.size.CropToAspectRatio`, but uses ``position="center"`` by default, which spreads the crop amounts equally over all image sides, while :class:`~imgaug.augmenters.size.CropToAspectRatio` by default spreads them randomly. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- aspect_ratio : number See :func:`CropToAspectRatio.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CenterCropToAspectRatio(2.0) Create an augmenter that crops each image until its aspect ratio is as close as possible to ``2.0`` (i.e. two times as many pixels along the x-axis than the y-axis). The rows to be cropped will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, aspect_ratio, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterCropToAspectRatio, self).__init__( aspect_ratio=aspect_ratio, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class PadToAspectRatio(PadToFixedSize): """Pad images until their width/height matches an aspect ratio. This augmenter adds either rows or columns until the image reaches the desired aspect ratio given in ``width / height``. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- aspect_ratio : number The desired aspect ratio, given as ``width/height``. E.g. a ratio of ``2.0`` denotes an image that is twice as wide as it is high. position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`PadToFixedSize.__init__`. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.PadToAspectRatio(2.0) Create an augmenter that pads each image until its aspect ratio is as close as possible to ``2.0`` (i.e. two times as many pixels along the x-axis than the y-axis). The rows to be padded will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, aspect_ratio, pad_mode="constant", pad_cval=0, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(PadToAspectRatio, self).__init__( width=None, height=None, pad_mode=pad_mode, pad_cval=pad_cval, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.aspect_ratio = aspect_ratio # Added in 0.4.0. def _draw_samples(self, batch, random_state): _sizes, pad_xs, pad_ys, pad_modes, pad_cvals = super( PadToAspectRatio, self )._draw_samples(batch, random_state) shapes = batch.get_rowwise_shapes() sizes = [] for shape in shapes: height, width = shape[0:2] paddings = compute_paddings_to_reach_aspect_ratio( shape, aspect_ratio=self.aspect_ratio) # TODO change that # note that these are not in the same order as shape tuples # in PadToFixedSize new_size = ( width + paddings[1] + paddings[3], height + paddings[0] + paddings[2] ) sizes.append(new_size) return sizes, pad_xs, pad_ys, pad_modes, pad_cvals # Added in 0.4.0. def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.aspect_ratio, self.pad_mode, self.pad_cval, self.position] class CenterPadToAspectRatio(PadToAspectRatio): """Pad images equally on all sides until H/W matches an aspect ratio. This is the same as :class:`~imgaug.augmenters.size.PadToAspectRatio`, but uses ``position="center"`` by default, which spreads the pad amounts equally over all image sides, while :class:`~imgaug.augmenters.size.PadToAspectRatio` by default spreads them randomly. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- aspect_ratio : number See :func:`PadToAspectRatio.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToAspectRatio.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToAspectRatio.__init__`. deterministic : bool, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.PadToAspectRatio(2.0) Create am augmenter that pads each image until its aspect ratio is as close as possible to ``2.0`` (i.e. two times as many pixels along the x-axis than the y-axis). The rows to be padded will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, aspect_ratio, pad_mode="constant", pad_cval=0, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterPadToAspectRatio, self).__init__( aspect_ratio=aspect_ratio, position="center", pad_mode=pad_mode, pad_cval=pad_cval, seed=seed, name=name, random_state=random_state, deterministic=deterministic) class CropToSquare(CropToAspectRatio): """Crop images until their width and height are identical. This is identical to :class:`~imgaug.augmenters.size.CropToAspectRatio` with ``aspect_ratio=1.0``. Images with axis sizes of ``0`` will not be altered. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`CropToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CropToSquare() Create an augmenter that crops each image until its square, i.e. height and width match. The rows to be cropped will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CropToSquare, self).__init__( aspect_ratio=1.0, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) class CenterCropToSquare(CropToSquare): """Crop images equally on all sides until their height/width are identical. In contrast to :class:`~imgaug.augmenters.size.CropToSquare`, this augmenter always tries to spread the columns/rows to remove equally over both sides of the respective axis to be cropped. :class:`~imgaug.augmenters.size.CropToAspectRatio` by default spreads the croppings randomly. This augmenter is identical to :class:`~imgaug.augmenters.size.CropToSquare` with ``position="center"``, and thereby the same as :class:`~imgaug.augmenters.size.CropToAspectRatio` with ``aspect_ratio=1.0, position="center"``. Images with axis sizes of ``0`` will not be altered. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.CropToFixedSize`. Parameters ---------- seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CenterCropToSquare() Create an augmenter that crops each image until its square, i.e. height and width match. The rows to be cropped will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterCropToSquare, self).__init__( position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class PadToSquare(PadToAspectRatio): """Pad images until their height and width are identical. This augmenter is identical to :class:`~imgaug.augmenters.size.PadToAspectRatio` with ``aspect_ratio=1.0``. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- position : {'uniform', 'normal', 'center', 'left-top', 'left-center', 'left-bottom', 'center-top', 'center-center', 'center-bottom', 'right-top', 'right-center', 'right-bottom'} or tuple of float or StochasticParameter or tuple of StochasticParameter, optional See :func:`PadToFixedSize.__init__`. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToFixedSize.__init__`. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.PadToSquare() Create an augmenter that pads each image until its square, i.e. height and width match. The rows to be padded will be spread *randomly* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, pad_mode="constant", pad_cval=0, position="uniform", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(PadToSquare, self).__init__( aspect_ratio=1.0, pad_mode=pad_mode, pad_cval=pad_cval, position=position, seed=seed, name=name, random_state=random_state, deterministic=deterministic) class CenterPadToSquare(PadToSquare): """Pad images equally on all sides until their height & width are identical. This is the same as :class:`~imgaug.augmenters.size.PadToSquare`, but uses ``position="center"`` by default, which spreads the pad amounts equally over all image sides, while :class:`~imgaug.augmenters.size.PadToSquare` by default spreads them randomly. This augmenter is thus also identical to :class:`~imgaug.augmenters.size.PadToAspectRatio` with ``aspect_ratio=1.0, position="center"``. Added in 0.4.0. **Supported dtypes**: See :class:`~imgaug.augmenters.size.PadToFixedSize`. Parameters ---------- name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. pad_mode : imgaug.ALL or str or list of str or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToAspectRatio.__init__`. pad_cval : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional See :func:`~imgaug.augmenters.size.PadToAspectRatio.__init__`. deterministic : bool, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.CenterPadToSquare() Create an augmenter that pads each image until its square, i.e. height and width match. The rows to be padded will be spread *equally* over the top and bottom sides (analogous for the left/right sides). """ # Added in 0.4.0. def __init__(self, pad_mode="constant", pad_cval=0, seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(CenterPadToSquare, self).__init__( pad_mode=pad_mode, pad_cval=pad_cval, position="center", seed=seed, name=name, random_state=random_state, deterministic=deterministic) class KeepSizeByResize(meta.Augmenter): """Resize images back to their input sizes after applying child augmenters. Combining this with e.g. a cropping augmenter as the child will lead to images being resized back to the input size after the crop operation was applied. Some augmenters have a ``keep_size`` argument that achieves the same goal (if set to ``True``), though this augmenter offers control over the interpolation mode and which augmentables to resize (images, heatmaps, segmentation maps). **Supported dtypes**: See :func:`~imgaug.imgaug.imresize_many_images`. Parameters ---------- children : Augmenter or list of imgaug.augmenters.meta.Augmenter or None, optional One or more augmenters to apply to images. These augmenters may change the image size. interpolation : KeepSizeByResize.NO_RESIZE or {'nearest', 'linear', 'area', 'cubic'} or {cv2.INTER_NEAREST, cv2.INTER_LINEAR, cv2.INTER_AREA, cv2.INTER_CUBIC} or list of str or list of int or StochasticParameter, optional The interpolation mode to use when resizing images. Can take any value that :func:`~imgaug.imgaug.imresize_single_image` accepts, e.g. ``cubic``. * If this is ``KeepSizeByResize.NO_RESIZE`` then images will not be resized. * If this is a single ``str``, it is expected to have one of the following values: ``nearest``, ``linear``, ``area``, ``cubic``. * If this is a single integer, it is expected to have a value identical to one of: ``cv2.INTER_NEAREST``, ``cv2.INTER_LINEAR``, ``cv2.INTER_AREA``, ``cv2.INTER_CUBIC``. * If this is a ``list`` of ``str`` or ``int``, it is expected that each ``str``/``int`` is one of the above mentioned valid ones. A random one of these values will be sampled per image. * If this is a ``StochasticParameter``, it will be queried once per call to ``_augment_images()`` and must return ``N`` ``str`` s or ``int`` s (matching the above mentioned ones) for ``N`` images. interpolation_heatmaps : KeepSizeByResize.SAME_AS_IMAGES or KeepSizeByResize.NO_RESIZE or {'nearest', 'linear', 'area', 'cubic'} or {cv2.INTER_NEAREST, cv2.INTER_LINEAR, cv2.INTER_AREA, cv2.INTER_CUBIC} or list of str or list of int or StochasticParameter, optional The interpolation mode to use when resizing heatmaps. Meaning and valid values are similar to `interpolation`. This parameter may also take the value ``KeepSizeByResize.SAME_AS_IMAGES``, which will lead to copying the interpolation modes used for the corresponding images. The value may also be returned on a per-image basis if `interpolation_heatmaps` is provided as a ``StochasticParameter`` or may be one possible value if it is provided as a ``list`` of ``str``. interpolation_segmaps : KeepSizeByResize.SAME_AS_IMAGES or KeepSizeByResize.NO_RESIZE or {'nearest', 'linear', 'area', 'cubic'} or {cv2.INTER_NEAREST, cv2.INTER_LINEAR, cv2.INTER_AREA, cv2.INTER_CUBIC} or list of str or list of int or StochasticParameter, optional The interpolation mode to use when resizing segmentation maps. Similar to `interpolation_heatmaps`. **Note**: For segmentation maps, only ``NO_RESIZE`` or nearest neighbour interpolation (i.e. ``nearest``) make sense in the vast majority of all cases. seed : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. name : None or str, optional See :func:`~imgaug.augmenters.meta.Augmenter.__init__`. random_state : None or int or imgaug.random.RNG or numpy.random.Generator or numpy.random.BitGenerator or numpy.random.SeedSequence or numpy.random.RandomState, optional Old name for parameter `seed`. Its usage will not yet cause a deprecation warning, but it is still recommended to use `seed` now. Outdated since 0.4.0. deterministic : bool, optional Deprecated since 0.4.0. See method ``to_deterministic()`` for an alternative and for details about what the "deterministic mode" actually does. Examples -------- >>> import imgaug.augmenters as iaa >>> aug = iaa.KeepSizeByResize( >>> iaa.Crop((20, 40), keep_size=False) >>> ) Apply random cropping to input images, then resize them back to their original input sizes. The resizing is done using this augmenter instead of the corresponding internal resizing operation in ``Crop``. >>> aug = iaa.KeepSizeByResize( >>> iaa.Crop((20, 40), keep_size=False), >>> interpolation="nearest" >>> ) Same as in the previous example, but images are now always resized using nearest neighbour interpolation. >>> aug = iaa.KeepSizeByResize( >>> iaa.Crop((20, 40), keep_size=False), >>> interpolation=["nearest", "cubic"], >>> interpolation_heatmaps=iaa.KeepSizeByResize.SAME_AS_IMAGES, >>> interpolation_segmaps=iaa.KeepSizeByResize.NO_RESIZE >>> ) Similar to the previous example, but images are now sometimes resized using linear interpolation and sometimes using nearest neighbour interpolation. Heatmaps are resized using the same interpolation as was used for the corresponding image. Segmentation maps are not resized and will therefore remain at their size after cropping. """ NO_RESIZE = "NO_RESIZE" SAME_AS_IMAGES = "SAME_AS_IMAGES" def __init__(self, children, interpolation="cubic", interpolation_heatmaps=SAME_AS_IMAGES, interpolation_segmaps="nearest", seed=None, name=None, random_state="deprecated", deterministic="deprecated"): super(KeepSizeByResize, self).__init__( seed=seed, name=name, random_state=random_state, deterministic=deterministic) self.children = children def _validate_param(val, allow_same_as_images): valid_ips_and_resize = ia.IMRESIZE_VALID_INTERPOLATIONS \ + [KeepSizeByResize.NO_RESIZE] if allow_same_as_images and val == self.SAME_AS_IMAGES: return self.SAME_AS_IMAGES if val in valid_ips_and_resize: return iap.Deterministic(val) if isinstance(val, list): assert len(val) > 0, ( "Expected a list of at least one interpolation method. " "Got an empty list.") valid_ips_here = valid_ips_and_resize if allow_same_as_images: valid_ips_here = valid_ips_here \ + [KeepSizeByResize.SAME_AS_IMAGES] only_valid_ips = all([ip in valid_ips_here for ip in val]) assert only_valid_ips, ( "Expected each interpolations to be one of '%s', got " "'%s'." % (str(valid_ips_here), str(val))) return iap.Choice(val) if isinstance(val, iap.StochasticParameter): return val raise Exception( "Expected interpolation to be one of '%s' or a list of " "these values or a StochasticParameter. Got type %s." % ( str(ia.IMRESIZE_VALID_INTERPOLATIONS), type(val))) self.children = meta.handle_children_list(children, self.name, "then") self.interpolation = _validate_param(interpolation, False) self.interpolation_heatmaps = _validate_param(interpolation_heatmaps, True) self.interpolation_segmaps = _validate_param(interpolation_segmaps, True) # Added in 0.4.0. def _augment_batch_(self, batch, random_state, parents, hooks): with batch.propagation_hooks_ctx(self, hooks, parents): images_were_array = None if batch.images is not None: images_were_array = ia.is_np_array(batch.images) shapes_orig = self._get_shapes(batch) samples = self._draw_samples(batch.nb_rows, random_state) batch = self.children.augment_batch_( batch, parents=parents + [self], hooks=hooks) if batch.images is not None: batch.images = self._keep_size_images( batch.images, shapes_orig["images"], images_were_array, samples) if batch.heatmaps is not None: # dont use shapes_orig["images"] because they might be None batch.heatmaps = self._keep_size_maps( batch.heatmaps, shapes_orig["heatmaps"], shapes_orig["heatmaps_arr"], samples[1]) if batch.segmentation_maps is not None: # dont use shapes_orig["images"] because they might be None batch.segmentation_maps = self._keep_size_maps( batch.segmentation_maps, shapes_orig["segmentation_maps"], shapes_orig["segmentation_maps_arr"], samples[2]) for augm_name in ["keypoints", "bounding_boxes", "polygons", "line_strings"]: augm_value = getattr(batch, augm_name) if augm_value is not None: func = functools.partial( self._keep_size_keypoints, shapes_orig=shapes_orig[augm_name], interpolations=samples[0]) cbaois = self._apply_to_cbaois_as_keypoints(augm_value, func) setattr(batch, augm_name, cbaois) return batch # Added in 0.4.0. @classmethod def _keep_size_images(cls, images, shapes_orig, images_were_array, samples): interpolations, _, _ = samples gen = zip(images, interpolations, shapes_orig) result = [] for image, interpolation, input_shape in gen: if interpolation == KeepSizeByResize.NO_RESIZE: result.append(image) else: result.append( ia.imresize_single_image(image, input_shape[0:2], interpolation)) if images_were_array: # note here that NO_RESIZE can have led to different shapes nb_shapes = len({image.shape for image in result}) if nb_shapes == 1: # images.dtype does not necessarily work anymore, children # might have turned 'images' into list result = np.array(result, dtype=result[0].dtype) return result # Added in 0.4.0. @classmethod def _keep_size_maps(cls, augmentables, shapes_orig_images, shapes_orig_arrs, interpolations): result = [] gen = zip(augmentables, interpolations, shapes_orig_arrs, shapes_orig_images) for augmentable, interpolation, arr_shape_orig, img_shape_orig in gen: if interpolation == "NO_RESIZE": result.append(augmentable) else: augmentable = augmentable.resize( arr_shape_orig[0:2], interpolation=interpolation) augmentable.shape = img_shape_orig result.append(augmentable) return result # Added in 0.4.0. @classmethod def _keep_size_keypoints(cls, kpsois_aug, shapes_orig, interpolations): result = [] gen = zip(kpsois_aug, interpolations, shapes_orig) for kpsoi_aug, interpolation, input_shape in gen: if interpolation == KeepSizeByResize.NO_RESIZE: result.append(kpsoi_aug) else: result.append(kpsoi_aug.on_(input_shape)) return result # Added in 0.4.0. @classmethod def _get_shapes(cls, batch): result = dict() for column in batch.columns: result[column.name] = [cell.shape for cell in column.value] if batch.heatmaps is not None: result["heatmaps_arr"] = [ cell.arr_0to1.shape for cell in batch.heatmaps] if batch.segmentation_maps is not None: result["segmentation_maps_arr"] = [ cell.arr.shape for cell in batch.segmentation_maps] return result def _draw_samples(self, nb_images, random_state): rngs = random_state.duplicate(3) interpolations = self.interpolation.draw_samples((nb_images,), random_state=rngs[0]) if self.interpolation_heatmaps == KeepSizeByResize.SAME_AS_IMAGES: interpolations_heatmaps = np.copy(interpolations) else: interpolations_heatmaps = self.interpolation_heatmaps.draw_samples( (nb_images,), random_state=rngs[1] ) # Note that `interpolations_heatmaps == self.SAME_AS_IMAGES` # works here only if the datatype of the array is such that it # may contain strings. It does not work properly for e.g. # integer arrays and will produce a single bool output, even # for arrays with more than one entry. same_as_imgs_idx = [ip == self.SAME_AS_IMAGES for ip in interpolations_heatmaps] interpolations_heatmaps[same_as_imgs_idx] = \ interpolations[same_as_imgs_idx] if self.interpolation_segmaps == KeepSizeByResize.SAME_AS_IMAGES: interpolations_segmaps = np.copy(interpolations) else: # TODO This used previously the same seed as the heatmaps part # leading to the same sampled values. Was that intentional? # Doesn't look like it should be that way. interpolations_segmaps = self.interpolation_segmaps.draw_samples( (nb_images,), random_state=rngs[2] ) # Note that `interpolations_heatmaps == self.SAME_AS_IMAGES` # works here only if the datatype of the array is such that it # may contain strings. It does not work properly for e.g. # integer arrays and will produce a single bool output, even # for arrays with more than one entry. same_as_imgs_idx = [ip == self.SAME_AS_IMAGES for ip in interpolations_segmaps] interpolations_segmaps[same_as_imgs_idx] = \ interpolations[same_as_imgs_idx] return interpolations, interpolations_heatmaps, interpolations_segmaps def _to_deterministic(self): aug = self.copy() aug.children = aug.children.to_deterministic() aug.deterministic = True aug.random_state = self.random_state.derive_rng_() return aug def get_parameters(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_parameters`.""" return [self.interpolation, self.interpolation_heatmaps] def get_children_lists(self): """See :func:`~imgaug.augmenters.meta.Augmenter.get_children_lists`.""" return [self.children] def __str__(self): pattern = ( "%s(" "interpolation=%s, " "interpolation_heatmaps=%s, " "name=%s, " "children=%s, " "deterministic=%s" ")") return pattern % ( self.__class__.__name__, self.interpolation, self.interpolation_heatmaps, self.name, self.children, self.deterministic)