Custom Object Detector implementation

yolo.py

+import cv2
+import numpy as np 
+import argparse
+import time
+import pandas as pd
+
+parser = argparse.ArgumentParser()
+parser.add_argument('--webcam', help="True/False", default=False)
+parser.add_argument('--play_video', help="Tue/False", default=False)
+parser.add_argument('--image', help="Tue/False", default=False)
+parser.add_argument('--video_path', help="Path of video file", default="Videos\DataCollection\ASD2.mp4")
+parser.add_argument('--image_path', help="Path of image to detect objects", default="Images\_000050.jpg")
+parser.add_argument('--verbose', help="To print statements", default=True)
+args = parser.parse_args()
+
+data = {'frame_number':[],'x_center':[], 'y_center':[]} #sr
+
+#Load yolo
+def load_yolo():
+	net = cv2.dnn.readNet("yolov3_custom_last_eyediapBall.weights", "yolov3_custom_eyediapball.cfg")
+	classes = []
+	with open("obj.names", "r") as f:
+		classes = [line.strip() for line in f.readlines()] 
+	
+	output_layers = [layer_name for layer_name in net.getUnconnectedOutLayersNames()]
+	colors = np.random.uniform(0, 255, size=(len(classes), 3))
+	return net, classes, colors, output_layers
+
+def load_image(img_path):
+	# image loading
+	img = cv2.imread(img_path)
+	img = cv2.resize(img, None, fx=0.4, fy=0.4)
+	height, width, channels = img.shape
+	return img, height, width, channels
+
+def start_webcam():
+	cap = cv2.VideoCapture(0)
+
+	return cap
+
+
+def display_blob(blob):
+	'''
+		Three images each for RED, GREEN, BLUE channel
+	'''
+	for b in blob:
+		for n, imgb in enumerate(b):
+			cv2.imshow(str(n), imgb)
+
+def detect_objects(img, net, outputLayers):			
+	blob = cv2.dnn.blobFromImage(img, scalefactor=0.00392, size=(320, 320), mean=(0, 0, 0), swapRB=True, crop=False)
+	net.setInput(blob)
+	outputs = net.forward(outputLayers)
+	return blob, outputs
+
+def get_box_dimensions(outputs, height, width):
+	boxes = []
+	confs = []
+	class_ids = []
+	for output in outputs:
+		for detect in output:
+			scores = detect[5:]
+			class_id = np.argmax(scores)
+			conf = scores[class_id]
+			if conf > 0.3:
+				center_x = int(detect[0] * width)
+				center_y = int(detect[1] * height)
+				w = int(detect[2] * width)
+				h = int(detect[3] * height)
+				x = int(center_x - w/2)
+				y = int(center_y - h / 2)
+				boxes.append([x, y, w, h])
+				confs.append(float(conf))
+				class_ids.append(class_id)
+
+				# print(center_x)
+				# print(center_y)
+	return boxes, confs, class_ids
+			
+def draw_labels(boxes, confs, colors, class_ids, classes, img): 
+	indexes = cv2.dnn.NMSBoxes(boxes, confs, 0.5, 0.4)
+	font = cv2.FONT_HERSHEY_PLAIN
+	# print(len(boxes))
+	temp_image = 0
+	print('Boxes:',boxes) 
+	print('Indexes:', indexes)
+	print('Classes:', classes, 'class_ids: ', class_ids)
+	if (len(boxes)== 0) or (len(indexes)== 0):
+		data['x_center'].append(np.nan)#sr
+		data['y_center'].append(np.nan)#sr
+
+
+	for i in range(len(boxes)):
+		if i in indexes:
+			x, y, w, h = boxes[i]
+			label = str(classes[class_ids[i]])
+			# print(colors)
+			color = colors[0]
+			cv2.rectangle(img, (x,y), (x+w, y+h), color, 2)
+			cv2.putText(img, label, (x, y - 5), font, 1, color, 1)
+
+			data['x_center'].append(x+(w/2))#sr
+			data['y_center'].append(y+(h/2))#sr
+			print(len(boxes))
+			
+			temp_image = img
+	return temp_image
+
+			# print('x+w/2', (x+(w/2)))
+			# print('y+h/2', (y+(h/2)))
+	# cv2.imshow("Image", img)
+
+	
+
+def image_detect(img_path): 
+	model, classes, colors, output_layers = load_yolo()
+	image, height, width, channels = load_image(img_path)
+	blob, outputs = detect_objects(image, model, output_layers)
+	boxes, confs, class_ids = get_box_dimensions(outputs, height, width)
+	draw_labels(boxes, confs, colors, class_ids, classes, image)
+	while True:
+		key = cv2.waitKey(1)
+		if key == 27:
+			break
+
+def webcam_detect():
+	model, classes, colors, output_layers = load_yolo()
+	cap = start_webcam()
+	while True:
+		_, frame = cap.read()
+		height, width, channels = frame.shape
+		blob, outputs = detect_objects(frame, model, output_layers)
+		boxes, confs, class_ids = get_box_dimensions(outputs, height, width)
+		draw_labels(boxes, confs, colors, class_ids, classes, frame)
+		key = cv2.waitKey(1)
+		if key == 27:
+			break
+	cap.release()
+
+#resizing arrays which are not same length	
+# def f(x):
+# 	vals = x[~x.isnull()].values
+# 	vals = np.resize(vals,len(x))
+# 	return vals
+
+def start_video(video_path):
+	model, classes, colors, output_layers = load_yolo()
+	cap = cv2.VideoCapture(video_path)
+	current_frame = 0 #sr
+
+	frame_width = int(cap.get(3))
+	frame_height = int(cap.get(4))
+   
+	size = (frame_width, frame_height)
+   
+	# Below VideoWriter object will create a frame of above defined The output is stored in 'filename.avi' file.
+	result = cv2.VideoWriter('Output Videos\DC\output_ASD2_FullyTrainedModel_er5.avi', 
+							cv2.VideoWriter_fourcc(*'MJPG'),
+                         30, size)
+
+	while cap.isOpened():
+		print(current_frame, len(data['x_center']), len(data['y_center']))
+
+		data['frame_number'].append(current_frame)
+		_, frame = cap.read()
+		try:
+			height, width, channels = frame.shape
+		except AttributeError:
+			print('NoneType frame reached!')
+			break
+		blob, outputs = detect_objects(frame, model, output_layers)
+		boxes, confs, class_ids = get_box_dimensions(outputs, height, width)
+		temp_image = draw_labels(boxes, confs, colors, class_ids, classes, frame)
+		key = cv2.waitKey(1)
+		if key == 27:
+			break
+		current_frame += 1#sr
+		result.write(temp_image)
+		# df_results = pd.DataFrame(data=data)
+		df_results = pd.DataFrame.from_dict(data=data, orient='index')
+		df_results = df_results.transpose()
+		print(df_results)
+		# df_results.to_csv('csv\_530COD_output_1_A_FT_S.csv', index=False,header=True, encoding='utf-8')
+		df_results.to_csv('csv\DC\Output_ASD2_FullyTrainedModel_er5.csv', index=False)
+		
+			
+	cap.release()
+
+
+if __name__ == '__main__':
+	webcam = args.webcam
+	video_play = args.play_video
+	image = args.image
+	if webcam:
+		if args.verbose:
+			print('---- Starting Web Cam object detection ----')
+		webcam_detect()
+	if video_play:
+		video_path = args.video_path
+		if args.verbose:
+			print('Opening '+video_path+" .... ")
+		start_video(video_path)
+	if image:
+		image_path = args.image_path
+		if args.verbose:
+			print("Opening "+image_path+" .... ")
+		image_detect(image_path)
+	
+
+	cv2.destroyAllWindows()
\ No newline at end of file