-
Notifications
You must be signed in to change notification settings - Fork 0
/
Detection_and_tracking(backup).py
208 lines (164 loc) · 6.82 KB
/
Detection_and_tracking(backup).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
import cv2
from darkflow.net.build import TFNet
import numpy as np
from time import sleep
import math
import datetime
import serial
from picamera.array import PiRGBArray
from picamera import PiCamera
options = {
'model': 'cfg/tiny-yolo-voc-1c.cfg',
'load': 3250,
'threshold': 0.2
}
ser = serial.Serial('/dev/ttyACM0', 57600)
tfnet = TFNet(options)
colors = [tuple(255 * np.random.rand(3)) for _ in range(10)]
file =open("data.txt","w")
# Initialise the camera setup
camera = PiCamera()
camera.resolution = (320, 320)
camera.saturation = 100
camera.brightness = 60
rawCapture = PiRGBArray(camera, size=(320, 320))
# allow the camera to warmup
time.sleep(5)
# Camera capture for the object detection
camera.capture(rawCapture, format='bgr', use_video_port = True)
image = rawCapture.array
cv2.imwrite('image.png', image)
cv2.waitKey(0)
camera.stop_preview()
camera.close()
frame = cv2.imread('image.png')
ret = 1
while True:
if ret:
results = tfnet.return_predict(frame)
for color, result in zip(colors, results):
tl = (result['topleft']['x'], result['topleft']['y'])
br = (result['bottomright']['x'], result['bottomright']['y'])
frame = cv2.rectangle(frame, tl, br, (0,0,255), 2)
w = result['bottomright']['x'] - result['topleft']['x']
h = result['bottomright']['y'] - result['topleft']['y']
cv2.imwrite('Rect.png',frame)
imCrop = frame[b:b+h, a:a+w]
# Thresholding to locate the center
hsv = cv2.cvtColor(imCrop, cv2.COLOR_BGR2HSV)
# Masking for the center
lower_red = np.array([30,110,110])
upper_red = np.array([80,255,255])
mask = cv2.inRange(hsv, lower_red, upper_red)
# Defining the morphological operations
kernel = np.ones((5,5),np.uint8)
erosion = cv2.erode(mask,kernel,iterations = 2)
kernel = np.ones((21,21),np.uint8)
dilate = cv2.dilate(erosion,kernel,iterations = 2)
# Finding the contours
M = cv2.moments(dilate)
cx = int(M['m10']/M['m00'])
cy = int(M['m01']/M['m00'])
cv2.circle(imCrop,(cx, cy), 2, (0,0,255), -1)
cy_n = b+cy
cx_n = a+cx
print("serial initialised")
sleep(5)
# Camera Initialisation for the video output
cam = PiCamera()
cam.resolution = (320,320)
cam.saturation = 100
cam.brightness = 60
rawCapture = PiRGBArray(cam, size=(320,320))
time.sleep(5)
# Parameters for LKOF
lk_params = dict( winSize = (15,15),
maxLevel = 4,
criteria = (cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, 10, 0.03))
cam.capture(rawCapture, format='bgr', use_video_port = True)
frame1 = rawCapture.array
old_gray = cv2.cvtColor(frame1, cv2.COLOR_BGR2GRAY)
rawCapture.truncate(0)
global point, point_selected, old_points
point = (cx_n, cy_n)
point_selected = True
old_points = np.array([[cx_n, cy_n]], dtype = np.float32)
for frame in cam.capture_continuous(rawCapture, format="bgr", use_video_port = True):
tm = datetime.datetime.utcnow().strftime("%H:%M:%S")
frame1 = frame.array
gray_frame = cv2.cvtColor(frame1, cv2.COLOR_BGR2GRAY)
if point_selected is True:
new_points, status, error = cv2.calcOpticalFlowPyrLK(old_gray, gray_frame, old_points, None, **lk_params)
old_gray = gray_frame.copy()
old_points = new_points
x, y = new_points.ravel()
# Bounding box size adjustment
if count<1:
im2, contours, hierarchy = cv2.findContours(dilate, 1, 2)
cnt = contours[0]
area = cv2.contourArea(cnt)
r = math.sqrt(area)
count=count+1
frame1 = cv2.rectangle(frame1, (int(x-(1*r)),int(y-(0.8*r))),(int(x+(1*r)),int(y+(0.8*r))), (0,0,255), 2)
# Extracting the new Bounding box
imCrop1 = frame1[int(y-(0.8*r)):int(y+(0.8*r)), int(x-(1*r)):int(x+(1*r))]
hsv1 = cv2.cvtColor(imCrop1, cv2.COLOR_BGR2HSV)
# Threshold value for red
lower_red = np.array([0,100,100])
upper_red = np.array([30,255,255])
mask_red = cv2.inRange(hsv1, lower_red, upper_red)
# Threshold value for green
lower_green = np.array([30,100,100])
upper_green = np.array([75,255,255])
mask_green = cv2.inRange(hsv1, lower_green, upper_green)
# Threshold value for blue
lower_blue = np.array([73,100,100])
upper_blue = np.array([120,255,255])
mask_blue = cv2.inRange(hsv1, lower_blue, upper_blue)
# Defining the morphological operations
kernel = np.ones((5,5),np.uint8)
#erosion_r = cv2.erode(mask_red,kernel,iterations = 2)
erosion_g = cv2.erode(mask_green,kernel,iterations = 2)
#erosion_b = cv2.erode(mask_blue,kernel,iterations = 2)
kernel = np.ones((21,21),np.uint8)
dilate_r = cv2.dilate(erosion_r,kernel,iterations = 2)
dilate_g = cv2.dilate(erosion_g,kernel,iterations = 2)
dilate_b = cv2.dilate(erosion_b,kernel,iterations = 2)
# Centroids of all the three targets
Mr = cv2.moments(dilate_r)
Mg = cv2.moments(dilate_g)
Mb = cv2.moments(dilate_b)
cx_g = int(Mg['m10']/Mg['m00'])
cy_g = int(Mg['m01']/Mg['m00'])
cx_r = int(Mr['m10']/Mr['m00'])
cy_r = int(Mr['m01']/Mr['m00'])
cx_b = int(Mb['m10']/Mb['m00'])
cy_b = int(Mb['m01']/Mb['m00'])
im2, contours, hierarchy = cv2.findContours(dilate_g, 1, 2)
cnt = contours[0]
area = cv2.contourArea(cnt)
r1 = math.sqrt(area)
r=r1
# Height and width of the bounding box
w=1.6*r
h=1*r
# Map the RoI coordinates wrt original frame
cx_g = x-(w/2)+cx_g
cy_g = y-(h/2)+cy_g
cx_r = x-(w/2)+cx_r
cy_r = y-(h/2)+cy_r
cx_b = x-(w/2)+cx_b
cy_b = y-(h/2)+cy_b
# file.write(str(x)+ ', ' + str(y) + ', ' + str(cx_g)+ ', ' + str(cy_g) + ', ' + str(cx_r)+ ', ' + str(cy_r) + ', ' + str(cx_b)+ ', ' + str(cy_b) + ', ' + str(w) + ', ' + str(h) + ', ' + str(pwm) + ', ' + str(tm) + ', ' +'\n')
cv2.imshow("Frame", frame1)
rawCapture.truncate(0)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
file.close()
cam1.release()
cv2.destroyAllWindows()
if cv2.waitKey(1) & 0xFF == ord('q'):
break
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cv2.destroyAllWindows()