File size: 10,451 Bytes
36c95ba
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
from typing import Tuple

import torch

from kornia.geometry.bbox import infer_bbox_shape, validate_bbox

from .imgwarp import get_perspective_transform, warp_affine

__all__ = [
    "crop_and_resize",
    "crop_by_boxes",
    "crop_by_transform_mat",
    "center_crop",
]


def crop_and_resize(
    tensor: torch.Tensor,
    boxes: torch.Tensor,
    size: Tuple[int, int],
    mode: str = 'bilinear',
    padding_mode: str = 'zeros',
    align_corners: bool = True,
) -> torch.Tensor:
    r"""Extract crops from 2D images (4D tensor) and resize given a bounding box.

    Args:
        tensor: the 2D image tensor with shape (B, C, H, W).
        boxes : a tensor containing the coordinates of the bounding boxes to be extracted.
            The tensor must have the shape of Bx4x2, where each box is defined in the following (clockwise)
            order: top-left, top-right, bottom-right and bottom-left. The coordinates must be in the x, y order.
            The coordinates would compose a rectangle with a shape of (N1, N2).
        size: a tuple with the height and width that will be
            used to resize the extracted patches.
        mode: interpolation mode to calculate output values
          ``'bilinear'`` | ``'nearest'``.
        padding_mode: padding mode for outside grid values
          ``'zeros'`` | ``'border'`` | 'reflection'.
        align_corners: mode for grid_generation.

    Returns:
        torch.Tensor: tensor containing the patches with shape BxCxN1xN2.

    Example:
        >>> input = torch.tensor([[[
        ...     [1., 2., 3., 4.],
        ...     [5., 6., 7., 8.],
        ...     [9., 10., 11., 12.],
        ...     [13., 14., 15., 16.],
        ... ]]])
        >>> boxes = torch.tensor([[
        ...     [1., 1.],
        ...     [2., 1.],
        ...     [2., 2.],
        ...     [1., 2.],
        ... ]])  # 1x4x2
        >>> crop_and_resize(input, boxes, (2, 2), mode='nearest', align_corners=True)
        tensor([[[[ 6.,  7.],
                  [10., 11.]]]])
    """
    if not isinstance(tensor, torch.Tensor):
        raise TypeError(f"Input tensor type is not a torch.Tensor. Got {type(tensor)}")

    if not isinstance(boxes, torch.Tensor):
        raise TypeError(f"Input boxes type is not a torch.Tensor. Got {type(boxes)}")

    if not isinstance(size, (tuple, list)) and len(size) == 2:
        raise ValueError(f"Input size must be a tuple/list of length 2. Got {size}")

    if len(tensor.shape) != 4:
        raise AssertionError(f"Only tensor with shape (B, C, H, W) supported. Got {tensor.shape}.")

    # unpack input data
    dst_h, dst_w = size

    # [x, y] origin
    # top-left, top-right, bottom-right, bottom-left
    points_src: torch.Tensor = boxes.to(tensor)

    # [x, y] destination
    # top-left, top-right, bottom-right, bottom-left
    points_dst: torch.Tensor = torch.tensor(
        [[[0, 0], [dst_w - 1, 0], [dst_w - 1, dst_h - 1], [0, dst_h - 1]]], device=tensor.device, dtype=tensor.dtype
    ).expand(points_src.shape[0], -1, -1)

    return crop_by_boxes(tensor, points_src, points_dst, mode, padding_mode, align_corners)


def center_crop(
    tensor: torch.Tensor,
    size: Tuple[int, int],
    mode: str = 'bilinear',
    padding_mode: str = 'zeros',
    align_corners: bool = True,
) -> torch.Tensor:
    r"""Crop the 2D images (4D tensor) from the center.

    Args:
        tensor: the 2D image tensor with shape (B, C, H, W).
        size: a tuple with the expected height and width
          of the output patch.
        mode: interpolation mode to calculate output values
          ``'bilinear'`` | ``'nearest'``.
        padding_mode: padding mode for outside grid values
          ``'zeros'`` | ``'border'`` | ``'reflection'``.
        align_corners: mode for grid_generation.

    Returns:
        the output tensor with patches.

    Examples:
        >>> input = torch.tensor([[[
        ...     [1., 2., 3., 4.],
        ...     [5., 6., 7., 8.],
        ...     [9., 10., 11., 12.],
        ...     [13., 14., 15., 16.],
        ...  ]]])
        >>> center_crop(input, (2, 4), mode='nearest', align_corners=True)
        tensor([[[[ 5.,  6.,  7.,  8.],
                  [ 9., 10., 11., 12.]]]])
    """
    if not isinstance(tensor, torch.Tensor):
        raise TypeError(f"Input tensor type is not a torch.Tensor. Got {type(tensor)}")

    if not isinstance(size, (tuple, list)) and len(size) == 2:
        raise ValueError(f"Input size must be a tuple/list of length 2. Got {size}")

    if len(tensor.shape) != 4:
        raise AssertionError(f"Only tensor with shape (B, C, H, W) supported. Got {tensor.shape}.")

    # unpack input sizes
    dst_h, dst_w = size
    src_h, src_w = tensor.shape[-2:]

    # compute start/end offsets
    dst_h_half: float = dst_h / 2
    dst_w_half: float = dst_w / 2
    src_h_half: float = src_h / 2
    src_w_half: float = src_w / 2

    start_x: float = src_w_half - dst_w_half
    start_y: float = src_h_half - dst_h_half

    end_x: float = start_x + dst_w - 1
    end_y: float = start_y + dst_h - 1

    # [y, x] origin
    # top-left, top-right, bottom-right, bottom-left
    points_src: torch.Tensor = torch.tensor(
        [[[start_x, start_y], [end_x, start_y], [end_x, end_y], [start_x, end_y]]],
        device=tensor.device,
        dtype=tensor.dtype,
    )

    # [y, x] destination
    # top-left, top-right, bottom-right, bottom-left
    points_dst: torch.Tensor = torch.tensor(
        [[[0, 0], [dst_w - 1, 0], [dst_w - 1, dst_h - 1], [0, dst_h - 1]]], device=tensor.device, dtype=tensor.dtype
    ).expand(points_src.shape[0], -1, -1)

    return crop_by_boxes(tensor, points_src, points_dst, mode, padding_mode, align_corners)


def crop_by_boxes(
    tensor: torch.Tensor,
    src_box: torch.Tensor,
    dst_box: torch.Tensor,
    mode: str = 'bilinear',
    padding_mode: str = 'zeros',
    align_corners: bool = True,
) -> torch.Tensor:
    """Perform crop transform on 2D images (4D tensor) given two bounding boxes.

    Given an input tensor, this function selected the interested areas by the provided bounding boxes (src_box).
    Then the selected areas would be fitted into the targeted bounding boxes (dst_box) by a perspective transformation.
    So far, the ragged tensor is not supported by PyTorch right now. This function hereby requires the bounding boxes
    in a batch must be rectangles with same width and height.

    Args:
        tensor: the 2D image tensor with shape (B, C, H, W).
        src_box: a tensor with shape (B, 4, 2) containing the coordinates of the bounding boxes
            to be extracted. The tensor must have the shape of Bx4x2, where each box is defined in the clockwise
            order: top-left, top-right, bottom-right and bottom-left. The coordinates must be in x, y order.
        dst_box: a tensor with shape (B, 4, 2) containing the coordinates of the bounding boxes
            to be placed. The tensor must have the shape of Bx4x2, where each box is defined in the clockwise
            order: top-left, top-right, bottom-right and bottom-left. The coordinates must be in x, y order.
        mode: interpolation mode to calculate output values
          ``'bilinear'`` | ``'nearest'``.
        padding_mode: padding mode for outside grid values
          ``'zeros'`` | ``'border'`` | ``'reflection'``.
        align_corners: mode for grid_generation.

    Returns:
        torch.Tensor: the output tensor with patches.

    Examples:
        >>> input = torch.arange(16, dtype=torch.float32).reshape((1, 1, 4, 4))
        >>> src_box = torch.tensor([[
        ...     [1., 1.],
        ...     [2., 1.],
        ...     [2., 2.],
        ...     [1., 2.],
        ... ]])  # 1x4x2
        >>> dst_box = torch.tensor([[
        ...     [0., 0.],
        ...     [1., 0.],
        ...     [1., 1.],
        ...     [0., 1.],
        ... ]])  # 1x4x2
        >>> crop_by_boxes(input, src_box, dst_box, align_corners=True)
        tensor([[[[ 5.0000,  6.0000],
                  [ 9.0000, 10.0000]]]])

    Note:
        If the src_box is smaller than dst_box, the following error will be thrown.
        RuntimeError: solve_cpu: For batch 0: U(2,2) is zero, singular U.
    """
    # TODO: improve this since might slow down the function
    validate_bbox(src_box)
    validate_bbox(dst_box)

    if len(tensor.shape) != 4:
        raise AssertionError(f"Only tensor with shape (B, C, H, W) supported. Got {tensor.shape}.")

    # compute transformation between points and warp
    # Note: Tensor.dtype must be float. "solve_cpu" not implemented for 'Long'
    dst_trans_src: torch.Tensor = get_perspective_transform(src_box.to(tensor), dst_box.to(tensor))

    bbox: Tuple[torch.Tensor, torch.Tensor] = infer_bbox_shape(dst_box)
    if not ((bbox[0] == bbox[0][0]).all() and (bbox[1] == bbox[1][0]).all()):
        raise AssertionError(
            f"Cropping height, width and depth must be exact same in a batch. "
            f"Got height {bbox[0]} and width {bbox[1]}."
        )

    h_out: int = int(bbox[0][0].item())
    w_out: int = int(bbox[1][0].item())

    return crop_by_transform_mat(
        tensor, dst_trans_src, (h_out, w_out), mode=mode, padding_mode=padding_mode, align_corners=align_corners
    )


def crop_by_transform_mat(
    tensor: torch.Tensor,
    transform: torch.Tensor,
    out_size: Tuple[int, int],
    mode: str = 'bilinear',
    padding_mode: str = 'zeros',
    align_corners: bool = True,
) -> torch.Tensor:
    """Perform crop transform on 2D images (4D tensor) given a perspective transformation matrix.

    Args:
        tensor: the 2D image tensor with shape (B, C, H, W).
        transform: a perspective transformation matrix with shape (B, 3, 3).
        out_size: size of the output image (height, width).
        mode: interpolation mode to calculate output values
          ``'bilinear'`` | ``'nearest'``.
        padding_mode (str): padding mode for outside grid values
          ``'zeros'`` | ``'border'`` | ``'reflection'``.
        align_corners: mode for grid_generation.

    Returns:
        the output tensor with patches.
    """
    # simulate broadcasting
    dst_trans_src = torch.as_tensor(transform.expand(tensor.shape[0], -1, -1), device=tensor.device, dtype=tensor.dtype)

    patches: torch.Tensor = warp_affine(
        tensor, dst_trans_src[:, :2, :], out_size, mode=mode, padding_mode=padding_mode, align_corners=align_corners
    )

    return patches