-
Notifications
You must be signed in to change notification settings - Fork 0
/
imgs2video.py
167 lines (150 loc) · 7.94 KB
/
imgs2video.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
import cv2
import os
import json
import numpy as np
from PIL import Image
import imgviz
import open3d as o3d
from open3d.examples.geometry.point_cloud_outlier_removal_statistical import display_inlier_outlier
from point import PointCloudHelper
# imgs_path = r"E:\Processed\dress-long-sleeve\dress-long-sleeve_001_20230731_134137\000673513312\color"
# for listx in os.listdir(imgs_path):
# print(listx)
# file_num = len([listx for listx in os.listdir(imgs_path)])
# img = cv2.imread(os.path.join(imgs_path,'0.png')) #读取第一张图片
# imgInfo = img.shape
# size = (imgInfo[1],imgInfo[0]) #获取图片宽高度信息
# print(file_num)
# fourcc = cv2.VideoWriter_fourcc(*'mp4v')
# videoWrite = cv2.VideoWriter('test.mp4',fourcc,10,size,True)# 根据图片的大小，创建写入对象 （文件名，支持的编码器，5帧，视频大小（图片大小））
#
# for i in range(1, file_num):
# fileName = str(i*3)+'.png' #循环读取所有的图片
# img = cv2.imread(os.path.join(imgs_path,fileName))
# print(img.shape)
# videoWrite.write(img)# 将图片写入所创建的视频对象
#
# videoWrite.release()
# print('end!')
def imgs2video(root_path=r"C:\Users\robotflow\Desktop\SAM\data0916\dress-long-sleeve_007_20230731_143236"):
camera_info_json_path = os.path.join(root_path, 'cameras_info.jsonl')
camera_info_json = json.load(open(camera_info_json_path, 'r'))
print(camera_info_json.keys())
for listx in os.listdir(root_path):
if listx.split("_")[0] == "SaveClip":
json_file_name = ""
for listx2 in os.listdir(os.path.join(root_path, listx)):
if listx2.split("_")[0] == "HandPose":
json_file_name = listx2
break
# print(json_file_name)
valid_json_path = os.path.join(root_path, listx, json_file_name)
valid_json = json.load(open(valid_json_path, 'r'))
valid_idexs = list(valid_json.keys())
print(valid_idexs)
file_num = len(valid_idexs)
img = cv2.imread(os.path.join(root_path, list(camera_info_json.keys())[0], "color", valid_idexs[0] + '.png'))
size = (img.shape[1], img.shape[0])
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
for camera_name in camera_info_json.keys():
# if(os.path.exists())
videoWriter = cv2.VideoWriter(os.path.join(root_path, listx, camera_name + ".mp4"), fourcc, 10, size, True)
for idex in valid_idexs:
img = np.zeros((size[1], size[0], 3), np.uint8)
if os.path.exists(os.path.join(root_path, camera_name, "color", idex + '.png')):
img = cv2.imread(os.path.join(root_path, camera_name, "color", idex + '.png'))
videoWriter.write(img)
videoWriter.release()
def mask2pc_test(masks_dir=r"C:\Users\robotflow\Desktop\SAM\Segment-and-Track-Anything-1.5\tracking_results\000673513312\000673513312_masks",
valid_hand_pose_json_path=r"C:\Users\robotflow\Desktop\SAM\data0916\dress-long-sleeve_007_20230731_143236\SaveClip_156_702\HandPose_156_702.json"):
masks_paths = os.listdir(masks_dir)
if not os.path.exists(os.path.join(masks_dir, "test")):
os.mkdir(os.path.join(masks_dir, "test"))
print(masks_paths)
valid_json = json.load(open(valid_hand_pose_json_path, 'r'))
valid_idexs = sorted(list(valid_json.keys()))
for i, mask_name in enumerate(masks_paths):
if not mask_name.endswith(".png"):
continue
# if not mask_name=="00100.png":
# continue
label = np.asarray(Image.open(os.path.join(masks_dir, mask_name)), dtype=np.int32)
# img = cv2.imread(os.path.join(masks_dir, mask_name))
print(label.shape)
print(label[250:300, 450:500])
label = label == 1
lbl_pil = Image.fromarray(label.astype(np.uint8), mode="P")
colormap = imgviz.label_colormap()
lbl_pil.putpalette(colormap.flatten())
lbl_pil.save(os.path.join(masks_dir, "test", mask_name))
# break
def mask2pc(clip_root=r"C:\Users\robotflow\Desktop\SAM\data0916\dress-long-sleeve_007_20230731_143236\SaveClip_156_702"):
if not os.path.exists(os.path.join(clip_root, "pcds")):
os.mkdir(os.path.join(clip_root, "pcds"))
root_path = os.path.dirname(clip_root)
camera_info_json_path = os.path.join(root_path, 'cameras_info.jsonl')
camera_info_json = json.load(open(camera_info_json_path, 'r'))
camera_extrinsics = {name: camera_info_json[name]['extrinsic'] for name in camera_info_json.keys()}
for name in camera_info_json.keys():
camera_info_json[name]["K"][0][0] = camera_info_json[name]["fxy"][0]
camera_info_json[name]["K"][1][1] = camera_info_json[name]["fxy"][1]
camera_info_json[name]["K"][0][2] = camera_info_json[name]["cxy"][0]
camera_info_json[name]["K"][1][2] = camera_info_json[name]["cxy"][1]
json_file_name = ""
for listx in os.listdir(clip_root):
if listx.split("_")[0] == "HandPose":
json_file_name = listx
break
if json_file_name == "":
print("HandPose***.json not found")
return
valid_json_path = os.path.join(clip_root, json_file_name)
valid_json = json.load(open(valid_json_path, 'r'))
valid_idexs = sorted(list(valid_json.keys()))
camera_masks = dict()
for camera_name in camera_info_json.keys():
if not os.path.exists(os.path.join(clip_root, camera_name + "_masks")):
print(camera_name + "'s masks dir not found")
return
img_filenames = []
for file in os.listdir(os.path.join(clip_root, camera_name + "_masks")):
if not file.endswith('.png'): continue
img_filenames.append(file)
if not len(valid_idexs) == len(img_filenames):
if len(valid_idexs) == len(img_filenames)+1:
img_filenames = ["00001.png"]+img_filenames
else:
print(len(valid_idexs), len(img_filenames))
return
camera_masks[camera_name] = img_filenames
print("valid clip and masks")
for i, idex in enumerate(valid_idexs):
print("processing frame: ", idex)
pcds = []
pcd_all = o3d.geometry.PointCloud()
for camera_name in camera_masks.keys():
img = cv2.imread(os.path.join(root_path, camera_name, "color", valid_idexs[i] + '.png'))
depth = cv2.imread(os.path.join(root_path, camera_name, "depth", valid_idexs[i] + '.png'), cv2.IMREAD_UNCHANGED)
mask = np.asarray(Image.open(os.path.join(clip_root, camera_name + "_masks", camera_masks[camera_name][i])), dtype=np.int32)
mask = mask == 1
depth = np.where(mask, depth, 0)
Camera_Tranform = np.eye(4)
Camera_Tranform[:3, :3] = np.array(camera_info_json[camera_name]["extrinsic"]["R"])
Camera_Tranform[:3, 3] = np.array(camera_info_json[camera_name]["extrinsic"]["T"])
pc = PointCloudHelper.rgbd2pc(img,
depth,
(3840 // 4, 2160 // 4, camera_info_json[camera_name]["K"]),
transform=Camera_Tranform,
enable_denoise=False)
pcds.append(pc)
pcd_all += pc
cl, ind = pcd_all.remove_statistical_outlier(nb_neighbors=50,
std_ratio=2.0)
output = pcd_all.select_by_index(ind)
#o3d.visualization.draw_geometries([output])
#display_inlier_outlier(pcd_all, ind)
o3d.io.write_point_cloud(os.path.join(clip_root, "pcds", valid_idexs[i] + '.ply'), output)
#break
if __name__ == "__main__":
imgs2video(r"\\tsclient\C\Users\robotflow\Desktop\SAM\data0916\dress-long-sleeve_013_20230731_142555")
#mask2pc(r"C:\Users\robotflow\Desktop\SAM\data0916\dress-long-sleeve_007_20230731_143236\SaveClip_156_702")