-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmain.py
148 lines (127 loc) · 6.74 KB
/
main.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
"""
+-+-+-+-+-+-+-+-+-+-+
|P|E|R|S|I|A|N|-|A|I|
+-+-+-+-+-+-+-+-+-+-+
"""
from ultralytics import YOLO
import cv2
import math
import time
import datetime
# Get the current timestamp for output names
timestamp = datetime.datetime.now().strftime("%Y%m%d%H%M%S")
# Define the file name with the timestamp
file_name = f'output_{timestamp}.jpg'
classnames = ['car', 'plate']
charclassnames = ['0','9','b','d','ein','ein','g','gh','h','n','s','1','malul','n','s','sad','t','ta','v','y','2'
,'3','4','5','6','7','8']
source = "assets/qq2.mp4"
#load YOLOv8 model
model_object = YOLO("weights/best.pt")
model_char = YOLO("weights/yolov8n_char_new.pt")
cap = cv2.VideoCapture(source)
# Define the output video properties
output_videoname = f'output_{timestamp}.mp4'
output_imagename = f'output_{timestamp}.jpg'
fourcc = cv2.VideoWriter_fourcc(*"mp4v")
fps = cap.get(cv2.CAP_PROP_FPS)
frame_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
if total_frames > 1:
video_writer = cv2.VideoWriter('output/' + output_videoname, fourcc, fps, (frame_width, frame_height))
#do inference for video
while cap.isOpened():
success, img = cap.read()
if success:
#detect objects with yolov8s model
tick = time.time()
output = model_object(img, show=False, conf=0.7, stream=True)
#extract bounding box and class names
for i in output:
bbox = i.boxes
for box in bbox:
x1, y1, x2, y2 = box.xyxy[0]
x1, y1, x2, y2 = int(x1), int(y1), int(x2), int(y2)
cv2.rectangle(img,(x1, y1), (x2, y2), (255, 0, 0), 3)
confs = math.ceil((box.conf[0]*100))/100
cls_names = int(box.cls[0])
if cls_names == 1:
cv2.putText(img, f'{confs}', (max(40, x2 + 5), max(40, y2 + 5)), fontFace=cv2.FONT_HERSHEY_TRIPLEX, fontScale=0.5, color=(0, 20, 255),thickness=1, lineType=cv2.LINE_AA)
elif cls_names == 0:
cv2.putText(img, f'{confs}', (max(40, x1), max(40, y1)), fontFace=cv2.FONT_HERSHEY_TRIPLEX, fontScale=0.6, color=(0, 20, 255),thickness=1, lineType=cv2.LINE_AA)
if cls_names == 1:
char_display = []
plate_img = img[ y1:y2, x1:x2]
plate_output = model_char(plate_img, conf=0.3)
tock_2 = time.time()
elapsed_time_2 = tock_2 - tick
bbox = plate_output[0].boxes.xyxy
cls = plate_output[0].boxes.cls
keys = cls.cpu().numpy().astype(int)
values = bbox[:, 0].cpu().numpy().astype(int)
dictionary = list(zip(keys, values))
sorted_list = sorted(dictionary, key=lambda x: x[1])
for i in sorted_list:
char_class = i[0]
char_display.append(charclassnames[char_class])
char_result ='Plate: ' + (''.join(char_display))
fps_text_2 = "FPS: {:.2f}".format(1/elapsed_time_2)
text_size, _ = cv2.getTextSize(fps_text_2, cv2.FONT_HERSHEY_SIMPLEX, 1, 2)
if len(char_display) == 8:
cv2.line(img, (max(40, x1 - 25 ), max(40, y1 - 10)), (x2 + 25 ,y1 - 10), (0, 0, 0), 20,lineType=cv2.LINE_AA)
cv2.putText(img, char_result , (max(40, x1 - 15), max(40, y1 - 5)), fontFace=cv2.FONT_HERSHEY_TRIPLEX, fontScale=0.5, color=(0, 201, 13),thickness=1, lineType=cv2.LINE_AA)
tock = time.time()
elapsed_time = tock - tick
fps_text = "FPS: {:.2f}".format(1/elapsed_time)
text_size, _ = cv2.getTextSize(fps_text, cv2.FONT_HERSHEY_SIMPLEX, 1, 2)
fps_text_loc = (frame_width - text_size[0] - 10, text_size[1] + 10)
cv2.putText(img ,
fps_text,fps_text_loc
, fontFace=cv2.FONT_HERSHEY_TRIPLEX, fontScale=1, color=(0, 201, 13),thickness=2, lineType=cv2.LINE_AA)
cv2.imshow('detection', img)
video_writer.write(img)
if cv2.waitKey(1) & 0xFF == ord("q"):
break
else:
break
cap.release()
cv2.destroyAllWindows()
else:
output = model_object(source, show=False, conf=0.75)
img = cv2.imread(source)
for i in output:
bbox = i.boxes
for box in bbox:
x1, y1, x2, y2 = box.xyxy[0]
x1, y1, x2, y2 = int(x1), int(y1), int(x2), int(y2)
cv2.rectangle(img,(x1, y1), (x2, y2), (255, 0, 0), 1)
confs = math.ceil((box.conf[0]*100))/100
cls_names = int(box.cls[0])
if cls_names == 1:
cv2.putText(img, f'{confs}', (max(40, x2 + 5), max(40, y2 + 5)), fontFace=cv2.FONT_HERSHEY_TRIPLEX, fontScale=0.5, color=(0, 201, 13),thickness=1, lineType=cv2.LINE_AA)
elif cls_names == 0:
cv2.putText(img, f'{confs}', (max(40, x1), max(40, y1)), fontFace=cv2.FONT_HERSHEY_TRIPLEX, fontScale=0.6, color=(0, 201, 13),thickness=1, lineType=cv2.LINE_AA)
if cls_names == 1:
char_display = []
plate_img = img[y1:y2, x1:x2]
plate_output = model_char(plate_img, conf=0.4)
bbox = plate_output[0].boxes.xyxy
cls = plate_output[0].boxes.cls
keys = cls.cpu().numpy().astype(int)
values = bbox[:, 0].cpu().numpy().astype(int)
dictionary = list(zip(keys, values))
sorted_list = sorted(dictionary, key=lambda x: x[1])
for i in sorted_list:
char_class = i[0]
char_display.append(plate_output[0].names[char_class])
char_display.append(charclassnames[char_class])
char_result = (''.join(char_display))
if len(char_display) == 8:
cv2.line(img, (max(40, x1 - 25 ), max(40, y1 - 10)), (x2 + 25 ,y1 - 10), (0, 0, 0), 20,lineType=cv2.LINE_AA)
cv2.putText(img, char_result , (max(40, x1 - 15), max(40, y1 - 5)), fontFace=cv2.FONT_HERSHEY_TRIPLEX, fontScale=0.5, color=(10, 50, 255),thickness=1, lineType=cv2.LINE_AA)
cv2.imshow('detection', img)
cv2.imwrite('output/' + output_imagename, img)
if cv2.waitKey(0) & 0xFF == ord("q"):
cap.release()
cv2.destroyAllWindows()