-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathwarp.py
executable file
·233 lines (184 loc) · 8.09 KB
/
warp.py
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
import tensorflow as tf
import cv2
import numpy as np
from scipy.ndimage.interpolation import map_coordinates as sp_map_coordinates
import matplotlib.tri as mtri
def tf_flatten(a):
"""Flatten tensor"""
return tf.reshape(a, [-1])
def tf_repeat(a, repeats, axis=0):
"""TensorFlow version of np.repeat for 1D"""
# https://github.com/tensorflow/tensorflow/issues/8521
assert len(a.get_shape()) == 1
a = tf.expand_dims(a, -1)
a = tf.tile(a, [1, repeats])
a = tf_flatten(a)
return a
def tf_repeat_2d(a, repeats):
"""Tensorflow version of np.repeat for 2D"""
assert len(a.get_shape()) == 2
a = tf.expand_dims(a, 0)
a = tf.tile(a, [repeats, 1, 1])
return a
def tf_map_coordinates(input, coords, order=1):
"""Tensorflow verion of scipy.ndimage.map_coordinates
Note that coords is transposed and only 2D is supported
Parameters
----------
input : tf.Tensor. shape = (s, s)
coords : tf.Tensor. shape = (n_points, 2)
"""
assert order == 1
coords_lt = tf.cast(tf.floor(coords), 'int32')
coords_rb = tf.cast(tf.ceil(coords), 'int32')
coords_lb = tf.stack([coords_lt[:, 0], coords_rb[:, 1]], axis=1)
coords_rt = tf.stack([coords_rb[:, 0], coords_lt[:, 1]], axis=1)
vals_lt = tf.gather_nd(input, coords_lt)
vals_rb = tf.gather_nd(input, coords_rb)
vals_lb = tf.gather_nd(input, coords_lb)
vals_rt = tf.gather_nd(input, coords_rt)
coords_offset_lt = coords - tf.cast(coords_lt, 'float32')
vals_t = vals_lt + (vals_rt - vals_lt) * coords_offset_lt[:, 0]
vals_b = vals_lb + (vals_rb - vals_lb) * coords_offset_lt[:, 0]
mapped_vals = vals_t + (vals_b - vals_t) * coords_offset_lt[:, 1]
return mapped_vals
def sp_batch_map_coordinates(inputs, coords):
"""Reference implementation for batch_map_coordinates"""
coords = coords.clip(0, inputs.shape[1] - 1)
mapped_vals = np.array([
sp_map_coordinates(input, coord.T, mode='nearest', order=1)
for input, coord in zip(inputs, coords)
])
return mapped_vals
def tf_batch_map_coordinates(_input, coords, order=1):
"""Batch version of tf_map_coordinates
Only supports 2D feature maps
Parameters
----------
input : tf.Tensor. shape = (b, s, s)
coords : tf.Tensor. shape = (b, n_points, 2)
"""
input_shape = tf.shape(_input)
batch_size = input_shape[0]
input_size = input_shape[1]
n_coords = tf.shape(coords)[1]
coords = tf.clip_by_value(coords, 0, tf.cast(input_size, 'float32') - 1)
coords_lt = tf.cast(tf.math.floor(coords), 'int32')
coords_rb = tf.cast(tf.math.ceil(coords), 'int32')
coords_lb = tf.stack([coords_lt[..., 0], coords_rb[..., 1]], axis=-1)
coords_rt = tf.stack([coords_rb[..., 0], coords_lt[..., 1]], axis=-1)
idx = tf_repeat(tf.range(batch_size), n_coords)
def _get_vals_by_coords(__input, coords):
indices = tf.stack([
idx, tf_flatten(coords[..., 0]), tf_flatten(coords[..., 1])
], axis=-1)
vals = tf.gather_nd(__input, indices)
vals = tf.reshape(vals, (batch_size, n_coords, __input.shape[3]))
return vals
vals_lt = _get_vals_by_coords(_input, coords_lt)
vals_rb = _get_vals_by_coords(_input, coords_rb)
vals_lb = _get_vals_by_coords(_input, coords_lb)
vals_rt = _get_vals_by_coords(_input, coords_rt)
coords_offset_lt = coords - tf.cast(coords_lt, 'float32')
offset_0 =coords_offset_lt[..., 0]
offset_1 =coords_offset_lt[..., 1]
offset_0 = tf.reshape(offset_0, [offset_0.shape[0], offset_0.shape[1], 1])
offset_1 = tf.reshape(offset_1, [offset_1.shape[0], offset_1.shape[1], 1])
vals_t = vals_lt + (vals_rt - vals_lt) * offset_0
vals_b = vals_lb + (vals_rb - vals_lb) * offset_0
mapped_vals = vals_t + (vals_b - vals_t) * offset_1
return mapped_vals
def sp_batch_map_offsets(input, offsets):
"""Reference implementation for tf_batch_map_offsets"""
batch_size = input.shape[0]
input_size = input.shape[1]
offsets = offsets.reshape(batch_size, -1, 2)
grid = np.stack(np.mgrid[:input_size, :input_size], -1).reshape(-1, 2)
grid = np.repeat([grid], batch_size, axis=0)
coords = offsets + grid
coords = coords.clip(0, input_size - 1)
mapped_vals = sp_batch_map_coordinates(input, coords)
return mapped_vals
def tf_batch_map_offsets(_input, offsets, order=1):
input_size = _input.shape[1]
#print(offsets, input_size)
offsets = tf.image.resize(offsets, [input_size, input_size]) * input_size
#offsets = offsets * imsize
offsets = offsets[:,:,:,0:2]
"""Batch map offsets into input
Parameters
---------
input : tf.Tensor. shape = (b, s, s)
offsets: tf.Tensor. shape = (b, s, s, 2)
"""
input_shape = tf.shape(_input)
batch_size = input_shape[0]
input_size = input_shape[1]
offsets = tf.reshape(offsets, (batch_size, -1, 2))
grid = tf.meshgrid(
tf.range(input_size), tf.range(input_size), indexing='ij'
)
grid = tf.stack(grid, axis=-1)
grid = tf.cast(grid, 'float32')
grid = tf.reshape(grid, (-1, 2))
grid = tf_repeat_2d(grid, batch_size)
coords = offsets + grid
mapped_vals = tf_batch_map_coordinates(_input, coords)
mapped_vals = tf.reshape(mapped_vals, (batch_size, input_size, input_size, -1))
return mapped_vals
'''def generate_offset_map(source, target):
offsetmap = []
anchor_pts = [[0,0],[0,255],[255,0],[255,255],
[0,127],[127,0],[255,127],[127,255],
[0,63],[0,191],[255,63],[255,191],
[63,0],[191,0],[63,255],[191,255]]
anchor_pts = np.asarray(anchor_pts)/ 255
xi, yi = np.meshgrid(np.linspace(0, 1, 256), np.linspace(0, 1, 256))
for i in range(source.shape[0]):
_source = np.concatenate([source[i,:,:], anchor_pts], axis=0).astype(np.float32)
_target = np.concatenate([target[i,:,:], anchor_pts], axis=0).astype(np.float32)
_offset = _source - _target
# interp2d
_triang = mtri.Triangulation(_target[:,0], _target[:,1])
_interpx = mtri.LinearTriInterpolator(_triang, _offset[:,0])
_interpy = mtri.LinearTriInterpolator(_triang, _offset[:,1])
_offsetmapx = _interpx(xi, yi)
_offsetmapy = _interpy(xi, yi)
_offsetmap = np.stack([_offsetmapy, _offsetmapx, _offsetmapx*0], axis=2)
offsetmap.append(_offsetmap)
offsetmap = np.stack(offsetmap, axis=0)
return offsetmap'''
def generate_offset_map(source, target, img_size):
anchor_pts = [[0,0],[0,255],[255,0],[255,255],
[0,127],[127,0],[255,127],[127,255],
[0,63],[0,191],[255,63],[255,191],
[63,0],[191,0],[63,255],[191,255]]
anchor_pts = np.asarray(anchor_pts)/ 255
xi, yi = np.meshgrid(np.linspace(0, 1, img_size), np.linspace(0, 1, img_size))
_source = np.concatenate([source, anchor_pts], axis=0).astype(np.float32)
_target = np.concatenate([target, anchor_pts], axis=0).astype(np.float32)
_offset = _source - _target
# interp2d
_triang = mtri.Triangulation(_target[:,0], _target[:,1])
_interpx = mtri.LinearTriInterpolator(_triang, _offset[:,0])
_interpy = mtri.LinearTriInterpolator(_triang, _offset[:,1])
_offsetmapx = _interpx(xi, yi)
_offsetmapy = _interpy(xi, yi)
offsetmap = np.stack([_offsetmapy, _offsetmapx, _offsetmapx*0], axis=2)
return offsetmap
def generate_uv_map(source, uv, img_size):
xi, yi = np.meshgrid(np.linspace(0, 1, img_size), np.linspace(0, 1, img_size))
# interp2d
_triang = mtri.Triangulation(source[:,0], source[:,1])
#_interpx = mtri.CubicTriInterpolator(_triang, uv[:,0])
#_interpy = mtri.CubicTriInterpolator(_triang, uv[:,1])
#_interpz = mtri.CubicTriInterpolator(_triang, uv[:,2])
_interpx = mtri.LinearTriInterpolator(_triang, uv[:,0])
_interpy = mtri.LinearTriInterpolator(_triang, uv[:,1])
_interpz = mtri.LinearTriInterpolator(_triang, uv[:,2])
_offsetmapx = _interpx(xi, yi)
_offsetmapy = _interpy(xi, yi)
_offsetmapz = _interpz(xi, yi)
offsetmap = np.stack([_offsetmapy, _offsetmapx, _offsetmapz], axis=2)
offsetmap = np.nan_to_num(offsetmap)
return offsetmap