-
Notifications
You must be signed in to change notification settings - Fork 0
/
mono_dataset.py
264 lines (224 loc) · 10.4 KB
/
mono_dataset.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
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
# Copyright Niantic 2019. Patent Pending. All rights reserved.
#
# This software is licensed under the terms of the Monodepth2 licence
# which allows for non-commercial use only, the full terms of which are made
# available in the LICENSE file.
from __future__ import absolute_import, division, print_function
import os
import random
import numpy as np
import copy
import PIL
from PIL import Image # using pillow-simd for increased speed
import torch
import torch.utils.data as data
from torchvision import transforms
from path import Path
import cv2
import time
def pil_loader(path):
# open path as file to avoid ResourceWarning
# (https://github.com/python-pillow/Pillow/issues/835)
with open(path, 'rb') as f:
with Image.open(f) as img:
return img.convert('RGB')
class MonoDataset(data.Dataset):
"""Superclass for monocular dataloaders
Args:
data_path
filenames
height
width
frame_idxs
num_scales
is_train
img_ext
"""
def __init__(self,
root_path,
height,
width,
frame_idxs,
num_scales,
with_pose ,
is_train=False):
super(MonoDataset, self).__init__()
self.root_path = Path(root_path)
self.height = height
self.width = width
self.frame_idxs = frame_idxs
self.num_scales = num_scales
self.interp = Image.ANTIALIAS
self.is_train = is_train
self.with_pose = with_pose
self.scene_list_path = self.root_path / 'train.txt' if is_train else self.root_path/ 'val.txt'
self.to_tensor = transforms.ToTensor()
self.scenes = [Path(folder[:-1]) for folder in open(self.scene_list_path)]
self.loader = pil_loader
self.to_tensor = transforms.ToTensor()
self.samples = self.crawl_folders(3)
# We need to specify augmentations differently in newer versions of torchvision.
# We first try the newer tuple version; if this fails we fall back to scalars
try:
self.brightness = (0.8, 1.2)
self.contrast = (0.8, 1.2)
self.saturation = (0.8, 1.2)
self.hue = (-0.1, 0.1)
transforms.ColorJitter.get_params(
self.brightness, self.contrast, self.saturation, self.hue)
except TypeError:
self.brightness = 0.2
self.contrast = 0.2
self.saturation = 0.2
self.hue = 0.1
self.resize = {}
for i in range(self.num_scales):
s = 2 ** i
self.resize[i] = transforms.Resize((self.height // s, self.width // s),
interpolation=self.interp)
self.load_depth = False
def preprocess(self, inputs, color_aug):
"""Resize colour images to the required scales and augment if required
We create the color_aug object in advance and apply the same augmentation to all
images in this item. This ensures that all images input to the pose network receive the
same augmentation.
"""
for k in list(inputs):
frame = inputs[k]
if "color" in k:
n, im, i = k
for i in range(self.num_scales):
inputs[(n, im, i)] = self.resize[i](inputs[(n, im, i - 1)])
inputs[(n,im,i)].save()
for k in list(inputs):
f = inputs[k]
if "color" in k:
n, im, i = k
inputs[(n, im, i)] = self.to_tensor(f)
inputs[(n + "_aug", im, i)] = self.to_tensor(color_aug(f))
def __len__(self):
return len(self.samples)
def __getitem__(self, index):
"""Returns a single training item from the dataset as a dictionary.
Values correspond to torch tensors.
Keys in the dictionary are either strings or tuples:
("color", <frame_id>, <scale>) for raw colour images,
("color_aug", <frame_id>, <scale>) for augmented colour images,
("K", scale) or ("inv_K", scale) for camera intrinsics,
"stereo_T" for camera extrinsics, and
"depth_gt" for ground truth depth maps.
<frame_id> is either:
an integer (e.g. 0, -1, or 1) representing the temporal step relative to 'index',
or
"s" for the opposite image in the stereo pair.
<scale> is an integer representing the scale of the image relative to the fullsize image:
-1 images at native resolution as loaded from disk
0 images resized to (self.width, self.height )
1 images resized to (self.width // 2, self.height // 2)
2 images resized to (self.width // 4, self.height // 4)
3 images resized to (self.width // 8, self.height // 8)
"""
inputs = {}
sample = self.samples[index].copy()
do_color_aug = self.is_train and random.random()>0.5
do_flip = self.is_train and random.random() > 0.5
for i in self.frame_idxs:
if i == 0:
inputs[("color", i, -1)] = self.get_color(sample['tgt_left'],do_flip)
if i == "s":
inputs[("color", i, -1)] = self.get_color(sample['tgt_right'],do_flip)
if i == -1:
inputs[("color", i, -1)] = self.get_color(sample['ref_left'][0],do_flip)
if i == 1:
inputs[("color", i, -1)] = self.get_color(sample['ref_left'][1],do_flip)
# adjusting intrinsics to match each scale in the pyramid
for scale in range(self.num_scales):
K = sample['intrinsics']
zeros = np.zeros((3,1))
K = np.c_[K,zeros]
zeros = np.array([0,0,0,1]).reshape(1,4)
K = np.r_[K,zeros].astype(np.float32)
K[0, :] /= 1280
K[1, :] /= 1024
K[0, :] *= self.width // (2 ** scale)
K[1, :] *= self.height // (2 ** scale)
inv_K = np.linalg.pinv(K)
inputs[("K", scale)] = torch.from_numpy(K)
inputs[("inv_K", scale)] = torch.from_numpy(inv_K)
if do_color_aug:
color_aug = transforms.ColorJitter.get_params(
self.brightness, self.contrast, self.saturation, self.hue)
else:
color_aug = (lambda x: x)
self.preprocess(inputs, color_aug)
for i in self.frame_idxs:
del inputs[("color", i, -1)]
del inputs[("color_aug", i, -1)]
if self.load_depth:
depth_gt = self.get_depth(sample['left_depth'], do_flip)
depth_gt = torch.from_numpy(depth_gt).float()
inputs["depth_gt"] = depth_gt
if "s" in self.frame_idxs:
factor = 0.1
if sample['LR_Distance'] > 4.36:
factor = 0.105
stereo_T = np.eye(4, dtype=np.float32)
baseline_sign = -1 if do_flip else 1
side_sign = -1
# if side == "l" else 1
stereo_T[0, 3] = side_sign * baseline_sign * factor
inputs["stereo_T"] = torch.from_numpy(stereo_T)
return inputs
def crawl_folders(self, sequence_length):
sequence_set = []
demi_length = (sequence_length - 1) // 2 # demi_length = 1
shifts = list(range(-demi_length, demi_length + 1)) # (-1,1)
shifts.pop(demi_length) # 除去0,前后帧的索引
for scene in self.scenes:
intrinsics = np.load(scene / 'rec_intrinsic.npy').astype(np.float32).reshape((3, 3))
imgs_left = sorted(Path(scene / 'data/rec_rgb').glob('Left*'))
imgs_right = sorted(Path(scene / 'data/rec_rgb').glob('Right*'))
current_depth = []
for img in imgs_left:
d = scene / 'data' / 'rec_gt_with_mask' / (
img.name[0:5] + 'gt' + img.name[10:16] + '.tiff')
assert (d.isfile()), "depth file {} not found".format(str(d))
current_depth.append(d)
LR_Matrix_path = scene / 'endoscope_calibration.yaml'
LR_Matrix = cv2.FileStorage(LR_Matrix_path, cv2.FileStorage_READ)
T = np.array(LR_Matrix.getNode('T').mat()).astype(float)
T = np.linalg.norm(T)
if len(imgs_left) < sequence_length:
continue
for i in range(demi_length, len(imgs_left) - demi_length): # 确保待处理的图片,每一帧都有前后帧
sample = {'intrinsics': intrinsics, 'tgt_left': imgs_left[i], 'tgt_right': imgs_right[i],
'ref_left': [], 'ref_right': [],'left_depth': current_depth[i],'LR_Distance': T}
if self.with_pose:
sample['tgt_json'] = []
sample['refs_jsons'] = []
name = imgs_left[i].name
number = Path(self.getting_number_from_path(name))
json_path = scene / 'data/frame_data/frame_data' + number + '.json'
sample['tgt_json'].append(json_path)
for j in shifts:
sample['ref_left'].append(imgs_left[i + j])
sample['ref_right'].append(imgs_right[i + j])
if self.with_pose:
name = imgs_left[i + j].name
number = Path(self.getting_number_from_path(name))
json_path = scene / 'data/frame_data/frame_data' + number + '.json'
sample['refs_jsons'].append(json_path)
sequence_set.append(sample) # 第i帧的DICT集合
random.shuffle(sequence_set)
return sequence_set
def load_as_float(self,path):
return cv2.imread(path).astype(np.float32)
def get_color(self, folder, do_flip):
raise NotImplementedError
def get_depth(self, folder,do_flip):
raise NotImplementedError
def getting_number_from_path(path):
name = path.name
number = filter(Path.isdigit, name)
number_str = ''.join(list(number))
return number_str