-
Notifications
You must be signed in to change notification settings - Fork 7
/
Copy pathview_transformation.py
179 lines (128 loc) · 6.84 KB
/
view_transformation.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
import cv2
import json
import tifffile as tiff
import numpy as np
import matplotlib.pyplot as plt
from os.path import join
def tiff_reader(tiff_file):
raw = tiff.imread(tiff_file)
scene_l = raw[:1024,:,:]
img_r = raw[1024:,:,:]
print("left point cloud shape:", scene_l.shape, scene_l.dtype)
return scene_l # return value is a cv mat
def camera_pose_reader(parameter_file):
with open(parameter_file) as para_json_file:
data = json.load(para_json_file)
camera_pose = data['camera-pose']
pose_transformation = np.array(camera_pose)
return pose_transformation # return value is an np array
def camera_param_reader(parameter_file):
with open(parameter_file) as para_json_file:
data = json.load(para_json_file)
camera_para = data['camera-calibration']
l_camera_matrix = np.array(camera_para['KL'])
#r_camera_matrix = np.array(camera_para['KR'])
l_dist_coeff = np.array(camera_para['DL'])
#r_dist_coeff = np.array(camera_para['DR'])
return l_camera_matrix, l_dist_coeff #, r_camera_matrix, r_dist_coeff
def view_transfer(camera_pose_1, camera_pose_2, point_cloud_1, image_1):
camera_pose_2_inv = np.linalg.inv(camera_pose_2)
transformation = np.transpose(np.dot(camera_pose_2_inv, camera_pose_1))
print(image_1.shape)
h,w = image_1.shape[:2]
homo_map = np.ones((h,w,1))
homo_point_cloud_1= np.concatenate((point_cloud_1, homo_map), axis=2)
transformed_point_cloud = np.tensordot(homo_point_cloud_1, transformation, axes=([2], [1]))
transformed_point_cloud = np.array(transformed_point_cloud[:,:,:3])
return transformed_point_cloud, image_1
def reprojection_img(transformed_point_cloud, image_original, l_camera_matrix, l_dist_coeff):
rvec = np.zeros((3,1))
tvec = np.zeros((3,1))
h,w = image_original.shape[:2]
print('transformed pc shape:', transformed_point_cloud.shape, transformed_point_cloud.dtype)
PC = transformed_point_cloud.reshape(-1, 3)
image_original_vector = image_original.reshape(-1,3)
repro_coor, jacobian = cv2.projectPoints(PC, rvec, tvec, l_camera_matrix, l_dist_coeff)
repro_coor = np.squeeze(repro_coor)
reproject_img = np.zeros([h,w,3])
for i in range(repro_coor.shape[0]):
x = int(repro_coor[i,0])
y = int(repro_coor[i,1])
if x < w and x > 0 and y < h and y > 0:
reproject_img[y,x,0] = image_original_vector[i,0]
reproject_img[y,x,1] = image_original_vector[i,1]
reproject_img[y,x,2] = image_original_vector[i,2]
count = 0
for i in range(reproject_img.shape[0]):
for j in range(reproject_img.shape[1]):
if reproject_img[i,j,2] == 0:
count += 1
print('hollow point:', count)
return reproject_img
def reprojection_depth(transformed_point_cloud, l_camera_matrix, l_dist_coeff):
rvec = np.zeros((3, 1))
tvec = np.zeros((3, 1))
h, w = transformed_point_cloud.shape[:2]
print('transformed pc shape:', transformed_point_cloud.shape, transformed_point_cloud.dtype)
PC = transformed_point_cloud.reshape(-1, 3)
repro_coor, jacobian = cv2.projectPoints(PC, rvec, tvec, l_camera_matrix, l_dist_coeff)
repro_coor = np.squeeze(repro_coor)
reproject_depth = np.zeros([h, w, 3])
for i in range(repro_coor.shape[0]):
x = int(repro_coor[i, 0])
y = int(repro_coor[i, 1])
if x < w and x > 0 and y < h and y > 0:
reproject_depth[y, x, 0] = PC[i, 0]
reproject_depth[y, x, 1] = PC[i, 1]
reproject_depth[y, x, 2] = PC[i, 2]
count = 0
for i in range(reproject_depth.shape[0]):
for j in range(reproject_depth.shape[1]):
if reproject_depth[i, j, 2] == 0:
count += 1
return reproject_depth
def get_scene_pair(pointcloud_filepath, camera_filepath, current_num,forward_step):
current_scene = tiff_reader(join(pointcloud_filepath, "scene_points%.6d.tiff" % current_num))
next_scene = tiff_reader(join(pointcloud_filepath, "scene_points%.6d.tiff" % (current_num + forward_step)))
current_camera_matrix, current_dist_coeff = camera_param_reader(join(camera_filepath, "frame_data%.6d.json" % current_num))
next_camera_matrix, next_dist_coeff = camera_param_reader(join(camera_filepath, "frame_data%.6d.json" % (current_num + forward_step)))
current_pose = camera_pose_reader(join(camera_filepath, "frame_data%.6d.json" % current_num))
next_pose = camera_pose_reader(join(camera_filepath, "frame_data%.6d.json" % (current_num + forward_step)))
return current_scene, next_scene, current_pose, next_pose, current_camera_matrix, current_dist_coeff, next_camera_matrix, next_dist_coeff
def get_img_pair(image_filepath, current_num, forward_step):
current_img = cv2.imread(join(image_filepath,'frame_data%.6d.png' % current_num))
next_img = cv2.imread(join(image_filepath, 'frame_data%.6d.png' % (current_num + forward_step)))
return current_img, next_img
def main():
rootpath = '/media/xiran_zhang/TOSHIBA EXT/MICCAI_SCARED/dataset3'
camera_filepath = join(rootpath, 'keyframe_1/data/frame_data')
pointcloud_filepath = join(rootpath, 'keyframe_1/data/scene_points')
img_filepath = join(rootpath, 'keyframe_1/data/left_finalpass')
current_num = 0
forward_step = 250
pc1, pc2, pose1, pose2, camera_para1, dist_coeff1, camera_para2, dist_coeff2 = get_scene_pair(pointcloud_filepath, camera_filepath, current_num, forward_step)
img1, img2 = get_img_pair(img_filepath, current_num, forward_step)
transformed_pc1, img_original = view_transfer(pose1, pose2, pc1, img1)
reproject_pointcloud = reprojection_depth(transformed_pc1, camera_para2, dist_coeff2)
tiff.imsave('current_transformed_pc.tiff', reproject_pointcloud.astype(np.float32))
cv2.imwrite('current_transformed_frame.png', img_original)
tiff.imsave('current_raw_pc.tiff', pc2.astype(np.float32))
cv2.imwrite('current_raw_frame.png',img2)
reproject_img2 = reprojection_img(reproject_pointcloud, img_original, camera_para2, dist_coeff2)
img2 = reprojection_img(pc2, img2, camera_para2, dist_coeff2)
cv2.imwrite('repro_img_2.png', reproject_img2)
cv2.imwrite('img_2.png', img2)
cv2.imwrite('diff.png', reproject_img2 / 2 + img2 / 2)
"""
transformed_cloud, img_original = view_transfer(pose1, pose2, pc1, img1)
reprojected_to_img2 = reprojection(transformed_cloud, img1, camera_para2, dist_coeff2)
cv2.imwrite('reprojected_to_2.png', reprojected_to_img2)
original_img1 = reprojection(pc1, img1, camera_para1, dist_coeff1)
cv2.imwrite('current_image.png', original_img1)
# reproject the original one
original_img2 = reprojection(pc2, img2, camera_para2, dist_coeff2)
cv2.imwrite('next_image.png', original_img2)
cv2.imwrite('difference.png', original_img2/2 + reprojected_to_img2/2)
"""
if __name__ == '__main__':
main()