-
Notifications
You must be signed in to change notification settings - Fork 2
/
app.py
275 lines (189 loc) · 9.74 KB
/
app.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
265
266
267
268
269
270
271
272
273
274
275
import streamlit as st
import numpy as np
from PIL import Image
from ultralytics import YOLO
import cv2
import easyocr
import pandas as pd
from util import set_background, write_csv
import uuid
import os
from streamlit_webrtc import webrtc_streamer
import av
set_background("./imgs/background.png")
folder_path = "./licenses_plates_imgs_detected/"
LICENSE_MODEL_DETECTION_DIR = './models/license_plate_detector.pt'
COCO_MODEL_DIR = "./models/yolov8n.pt"
reader = easyocr.Reader(['en'], gpu=False)
vehicles = [2]
header = st.container()
body = st.container()
coco_model = YOLO(COCO_MODEL_DIR)
license_plate_detector = YOLO(LICENSE_MODEL_DETECTION_DIR)
threshold = 0.15
state = "Uploader"
if "state" not in st.session_state :
st.session_state["state"] = "Uploader"
class VideoProcessor:
def recv(self, frame) :
img = frame.to_ndarray(format="bgr24")
img_to_an = img.copy()
img_to_an = cv2.cvtColor(img_to_an, cv2.COLOR_RGB2BGR)
license_detections = license_plate_detector(img_to_an)[0]
if len(license_detections.boxes.cls.tolist()) != 0 :
for license_plate in license_detections.boxes.data.tolist() :
x1, y1, x2, y2, score, class_id = license_plate
cv2.rectangle(img, (int(x1), int(y1)), (int(x2), int(y2)), (0, 255, 0), 3)
license_plate_crop = img[int(y1):int(y2), int(x1): int(x2), :]
license_plate_crop_gray = cv2.cvtColor(license_plate_crop, cv2.COLOR_BGR2GRAY)
license_plate_text, license_plate_text_score = read_license_plate(license_plate_crop_gray, img)
cv2.rectangle(img, (int(x1) - 40, int(y1) - 40), (int(x2) + 40, int(y1)), (255, 255, 255), cv2.FILLED)
cv2.putText(img,
str(license_plate_text),
(int((int(x1) + int(x2)) / 2) - 70, int(y1) - 10),
cv2.FONT_HERSHEY_SIMPLEX,
1,
(0, 0, 0),
3)
return av.VideoFrame.from_ndarray(img, format="bgr24")
def read_license_plate(license_plate_crop, img):
scores = 0
detections = reader.readtext(license_plate_crop)
width = img.shape[1]
height = img.shape[0]
if detections == [] :
return None, None
rectangle_size = license_plate_crop.shape[0]*license_plate_crop.shape[1]
plate = []
for result in detections:
length = np.sum(np.subtract(result[0][1], result[0][0]))
height = np.sum(np.subtract(result[0][2], result[0][1]))
if length*height / rectangle_size > 0.17:
bbox, text, score = result
text = result[1]
text = text.upper()
scores += score
plate.append(text)
if len(plate) != 0 :
return " ".join(plate), scores/len(plate)
else :
return " ".join(plate), 0
def model_prediction(img):
license_numbers = 0
results = {}
licenses_texts = []
img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
object_detections = coco_model(img)[0]
license_detections = license_plate_detector(img)[0]
if len(object_detections.boxes.cls.tolist()) != 0 :
for detection in object_detections.boxes.data.tolist() :
xcar1, ycar1, xcar2, ycar2, car_score, class_id = detection
if int(class_id) in vehicles :
cv2.rectangle(img, (int(xcar1), int(ycar1)), (int(xcar2), int(ycar2)), (0, 0, 255), 3)
else :
xcar1, ycar1, xcar2, ycar2 = 0, 0, 0, 0
car_score = 0
if len(license_detections.boxes.cls.tolist()) != 0 :
license_plate_crops_total = []
for license_plate in license_detections.boxes.data.tolist() :
x1, y1, x2, y2, score, class_id = license_plate
cv2.rectangle(img, (int(x1), int(y1)), (int(x2), int(y2)), (0, 255, 0), 3)
license_plate_crop = img[int(y1):int(y2), int(x1): int(x2), :]
img_name = '{}.jpg'.format(uuid.uuid1())
cv2.imwrite(os.path.join(folder_path, img_name), license_plate_crop)
license_plate_crop_gray = cv2.cvtColor(license_plate_crop, cv2.COLOR_BGR2GRAY)
license_plate_text, license_plate_text_score = read_license_plate(license_plate_crop_gray, img)
licenses_texts.append(license_plate_text)
if license_plate_text is not None and license_plate_text_score is not None :
license_plate_crops_total.append(license_plate_crop)
results[license_numbers] = {}
results[license_numbers][license_numbers] = {'car': {'bbox': [xcar1, ycar1, xcar2, ycar2], 'car_score': car_score},
'license_plate': {'bbox': [x1, y1, x2, y2],
'text': license_plate_text,
'bbox_score': score,
'text_score': license_plate_text_score}}
license_numbers+=1
write_csv(results, f"./csv_detections/detection_results.csv")
img_wth_box = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
return [img_wth_box, licenses_texts, license_plate_crops_total]
else :
img_wth_box = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
return [img_wth_box]
def change_state_uploader() :
st.session_state["state"] = "Uploader"
def change_state_camera() :
st.session_state["state"] = "Camera"
def change_state_live() :
st.session_state["state"] = "Live"
with header :
_, col1, _ = st.columns([0.2,1,0.1])
col1.title("💥 License Car Plate Detection 🚗")
_, col0, _ = st.columns([0.15,1,0.1])
col0.image("./imgs/test_background.jpg", width=500)
_, col4, _ = st.columns([0.1,1,0.2])
col4.subheader("Computer Vision Detection with YoloV8 🧪")
_, col, _ = st.columns([0.3,1,0.1])
col.image("./imgs/plate_test.jpg")
_, col5, _ = st.columns([0.05,1,0.1])
st.write("The differents models detect the car and the license plate in a given image, then extracts the info about the license using EasyOCR, and crop and save the license plate as a Image, with a CSV file with all the data. ")
with body :
_, col1, _ = st.columns([0.1,1,0.2])
col1.subheader("Check It-out the License Car Plate Detection Model 🔎!")
_, colb1, colb2, colb3 = st.columns([0.2, 0.7, 0.6, 1])
if colb1.button("Upload an Image", on_click=change_state_uploader) :
pass
elif colb2.button("Take a Photo", on_click=change_state_camera) :
pass
elif colb3.button("Live Detection", on_click=change_state_live) :
pass
if st.session_state["state"] == "Uploader" :
img = st.file_uploader("Upload a Car Image: ", type=["png", "jpg", "jpeg"])
elif st.session_state["state"] == "Camera" :
img = st.camera_input("Take a Photo: ")
elif st.session_state["state"] == "Live" :
webrtc_streamer(key="sample", video_processor_factory=VideoProcessor)
img = None
_, col2, _ = st.columns([0.3,1,0.2])
_, col5, _ = st.columns([0.8,1,0.2])
if img is not None:
image = np.array(Image.open(img))
col2.image(image, width=400)
if col5.button("Apply Detection"):
results = model_prediction(image)
if len(results) == 3 :
prediction, texts, license_plate_crop = results[0], results[1], results[2]
texts = [i for i in texts if i is not None]
if len(texts) == 1 and len(license_plate_crop) :
_, col3, _ = st.columns([0.4,1,0.2])
col3.header("Detection Results ✅:")
_, col4, _ = st.columns([0.1,1,0.1])
col4.image(prediction)
_, col9, _ = st.columns([0.4,1,0.2])
col9.header("License Cropped ✅:")
_, col10, _ = st.columns([0.3,1,0.1])
col10.image(license_plate_crop[0], width=350)
_, col11, _ = st.columns([0.45,1,0.55])
col11.success(f"License Number: {texts[0]}")
df = pd.read_csv(f"./csv_detections/detection_results.csv")
st.dataframe(df)
elif len(texts) > 1 and len(license_plate_crop) > 1 :
_, col3, _ = st.columns([0.4,1,0.2])
col3.header("Detection Results ✅:")
_, col4, _ = st.columns([0.1,1,0.1])
col4.image(prediction)
_, col9, _ = st.columns([0.4,1,0.2])
col9.header("License Cropped ✅:")
_, col10, _ = st.columns([0.3,1,0.1])
_, col11, _ = st.columns([0.45,1,0.55])
col7, col8 = st.columns([1,1])
for i in range(0, len(license_plate_crop)) :
col10.image(license_plate_crop[i], width=350)
col11.success(f"License Number {i}: {texts[i]}")
df = pd.read_csv(f"./csv_detections/detection_results.csv")
st.dataframe(df)
else :
prediction = results[0]
_, col3, _ = st.columns([0.4,1,0.2])
col3.header("Detection Results ✅:")
_, col4, _ = st.columns([0.3,1,0.1])
col4.image(prediction)