-
Notifications
You must be signed in to change notification settings - Fork 0
/
detect.py
152 lines (132 loc) · 5.58 KB
/
detect.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
import argparse
import os
import yaml
import glob
import torch
import torch.backends.cudnn as cudnn
import cv2
import time
import numpy as np
from tools import Logger
from model import YuDetectNet
import glob
parser = argparse.ArgumentParser(description='Face and Landmark Detection')
parser.add_argument('--config', '-c', type=str, help='config to test')
parser.add_argument('--model', '-m', type=str, help='model path to test')
parser.add_argument('-t', '--target', type=str, help='image/image folder/video path')
parser.add_argument('--confidence_threshold', type=float, help='confidence threshold to save result')
parser.add_argument('--device', type=str, default='cuda:0', help='cpu or cuda:0')
def arg_initial(args):
if args.config is not None:
assert os.path.exists(args.config)
workfolder = os.path.join("./workspace", os.path.basename(args.config)[:-5])
if not os.path.exists(workfolder):
os.makedirs(workfolder)
else:
workfolder = os.path.dirname(os.path.dirname(args.model))
cfg_list = glob.glob(os.path.join(workfolder, '*.yaml'))
assert len(cfg_list) == 1, 'Can`t comfire config file!'
args.config = cfg_list[0]
with open(args.config, mode='r', encoding='utf-8') as f:
cfg = yaml.safe_load(f)
save_dir = os.path.join(workfolder, 'detect_results')
if not os.path.exists(save_dir):
os.mkdir(save_dir)
cfg['test']['save_dir'] = save_dir
assert os.path.exists(args.target)
if args.confidence_threshold is not None:
cfg['test']['confidence_threshold'] = args.confidence_threshold
return cfg
def detect_image(net, img_path, cfg, device):
img = cv2.imread(img_path)
t0 = time.time()
dets = net.inference(img, scale=1., without_landmarks=False, device=device)
t1 = time.time()
dets = dets.cpu().numpy()
if len(dets) == 0:
print('Detect 0 taeget!')
return t1 - t0
scores = dets[:, -1]
dets = dets[:, :-1].astype(np.int32)
for det, score in zip(dets, scores):
img = cv2.rectangle(img, (det[0], det[1]), (det[2], det[3]), color=(0, 0, 255), thickness=1)
# img = cv2.putText(img, f"{score:4f}", (det[0], det[1] + 12), cv2.FONT_HERSHEY_DUPLEX, 0.5, (255, 255, 255))
ldms_num = int((det.shape[-1] - 4) / 2)
for i in range(ldms_num):
cv2.circle(img, (det[4 + 2 * i], det[5 + 2 * i]), 2, (255, 255, 0), thickness=1)
save_path = os.path.join(cfg['test']['save_dir'], os.path.basename(img_path))
cv2.imwrite(save_path, img)
print(f'Detect {0 if len(dets.shape) == 1 else dets.shape[0]} target, Save img to {save_path}')
return t1 - t0
def detect_video(net, video_path, cfg, device):
cap = cv2.VideoCapture(video_path)
size = (int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)), int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT)))
fps = cap.get(cv2.CAP_PROP_FPS)
save_path = os.path.join(cfg['test']['save_dir'], os.path.basename(video_path))
video_writer = cv2.VideoWriter(
save_path,
cv2.VideoWriter_fourcc('M','P','E','G'),
fps,
size
)
print('Detect video, please wait ...')
total_time = 0
num_frames = 0
while(True):
ret, frame = cap.read()
if ret:
t0 = time.time()
det = net.inference(frame, scale=1., without_landmarks=False, device=device)
total_time += time.time() - t0
num_frames += 1
det = det.cpu().numpy()
scores = det[:, -1]
det = det[:, :-1].astype(np.int32)
for det, score in zip(det, scores):
frame = cv2.rectangle(frame, (det[0], det[1]), (det[2], det[3]), color=(0, 0, 255), thickness=1)
frame = cv2.putText(frame, f"{score:4f}", (det[0], det[1] + 12), cv2.FONT_HERSHEY_DUPLEX, 0.5, (255, 255, 255))
for i in range(int((det.shape[-1] - 4) / 2)):
cv2.circle(frame, (det[4 + 2 * i], det[5 + 2 * i]), 2, (255, 255, 0), thickness=5)
video_writer.write(frame)
else:
break
cap.release()
video_writer.release()
print(f'Save video to {save_path}')
return total_time, num_frames
def main():
args = parser.parse_args()
cfg = arg_initial(args)
torch.set_grad_enabled(False)
print(f'Loading model from {args.model}')
net = YuDetectNet(cfg)
net.load_state_dict(torch.load(args.model))
net.eval()
net.to(args.device)
cudnn.benchmark = True
target = args.target
img_paths = []
if os.path.isdir(target):
img_paths.extend(glob.glob(os.path.join(target, '*.jpg')))
img_paths.extend(glob.glob(os.path.join(target, '*.jpeg')))
img_paths.extend(glob.glob(os.path.join(target, '*.png')))
img_paths.extend(glob.glob(os.path.join(target, '*.mp4')))
else:
img_paths.append(target)
print(f'{len(img_paths)} files to be detected...')
total_time = 0
num_frames = 0
for img_path in img_paths:
filename, tp = os.path.splitext(os.path.basename(img_path))
if tp.lower() in ('.jpg', '.jpeg', '.png'):
total_time += detect_image(net, img_path, cfg, device=args.device)
num_frames += 1
elif tp.lower() in ('.mp4'):
_total_time, _num_frames = detect_video(net, img_path, cfg, device=args.device)
total_time += _total_time
num_frames += _num_frames
else:
print(f'{img_path}: Unsupport file!')
print(f'Detect {num_frames} images and achieve {int(num_frames / total_time)} fps.')
if __name__ == "__main__":
main()