track object

ObjectTracking/trackObject.py

@@ -0,0 +1,79 @@
+import numpy as np
+import cv2
+
+capture = cv2.VideoCapture(0)
+
+# Color range of object in HSV
+color_range = [[0,0,0],[179,50,100]]
+
+points = []
+count = 0
+
+# Get first frame
+response, frame = capture.read()
+height, width = frame.shape[:2]
+
+while True:
+	response, frame = capture.read()
+	frame = cv2.flip(frame, 1)
+
+	# converting to hsv
+	hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
+
+	# threshold values between range
+	mask = cv2.inRange(hsv, np.array(color_range[0]), np.array(color_range[1]))
+
+	# Finding contours
+	_, contours, _ = cv2.findContours(mask.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+
+	# Computing centroids
+	centroids = int(height/2), int(width/2)
+	radius = 0
+
+	if len(contours) > 0:
+		# Find largest contour
+		biggest_contour = max(contours, key=cv2.contourArea)
+		(x,y), radius = cv2.minEnclosingCircle(biggest_contour)
+		M = cv2.moments(biggest_contour)
+		try:
+			a = int(M['m10']/M['m00'])
+			b = int(M['m01']/M['m00'])
+			centroids = (a, b) 
+		except:
+			a = int(height/2)
+			b = int(width/2)
+			centroids = a,b
+
+		# Threshold contours	
+		if radius > 25:
+			print (radius)
+			cv2.circle(frame, (int(x), int(y)), int(radius), (0,0,255), 2)
+			cv2.circle(frame, centroids, 5, (0,255,0), -1)
+
+	points.append(centroids)
+
+	# track points
+	if radius > 25:
+		print (radius)
+		for i in range(1, len(points)):
+			try:
+				cv2.line(frame, points[i-1], points[i], (0,255,0), 2)
+			except:
+				pass
+
+		count += 0					
+
+	else:
+		count += 1	
+
+		# Remove trail when no object in 10 frames
+		if count == 10:
+			points = []
+			count = 0		
+
+	cv2.imshow('Tracked Object', frame)
+	if cv2.waitKey(1) & 0xFF == ord('q'):
+		break
+
+capture.release()
+cv2.destroyAllWindows()