【PYTHON OPENCV】Object detection OpenCV DNN module using MobileNet SSD and caffe pre trained models

 """

Object detection using OpenCV DNN module using MobileNet-SSD and caffe pre-trained models MobileNetSSD_deploy.caffemodel: https://drive.google.com/uc?export=download&id=0B3gersZ2cHIxRm5PMWRoTkdHdHc MobileNetSSD_deploy.prototxt https://raw.githubusercontent.com/chuanqi305/MobileNet-SSD/daef68a6c2f5fbb8c88404266aa28180646d17e0/MobileNetSSD_deploy.prototxt """ # Import required packages: import cv2 from matplotlib import pyplot as plt def show_img_with_matplotlib(color_img, title, pos): """Shows an image using matplotlib capabilities""" img_RGB = color_img[:, :, ::-1] ax = plt.subplot(1, 1, pos) plt.imshow(img_RGB) plt.title(title) plt.axis('off') # Load the serialized caffe model from disk: net = cv2.dnn.readNetFromCaffe("MobileNetSSD_deploy.prototxt", "MobileNetSSD_deploy.caffemodel") # Load input image: image = cv2.imread("object_detection_test_image.png") # Prepare labels of the network (20 class labels + background): class_names = {0: 'background', 1: 'aeroplane', 2: 'bicycle', 3: 'bird', 4: 'boat', 5: 'bottle', 6: 'bus', 7: 'car', 8: 'cat', 9: 'chair', 10: 'cow', 11: 'diningtable', 12: 'dog', 13: 'horse', 14: 'motorbike', 15: 'person', 16: 'pottedplant', 17: 'sheep', 18: 'sofa', 19: 'train', 20: 'tvmonitor'} # Create the blob with a size of (300,300), mean subtraction values (127.5, 127.5, 127.5): # and also a scalefactor of 0.007843: blob = cv2.dnn.blobFromImage(image, 0.007843, (300, 300), (127.5, 127.5, 127.5)) print(blob.shape) # Feed the input blob to the network, perform inference and ghe the output: net.setInput(blob) detections = net.forward() # Get inference time: t, _ = net.getPerfProfile() print('Inference time: %.2f ms' % (t * 1000.0 / cv2.getTickFrequency())) # Size of frame resize (300x300) dim = 300 # Process all detections: for i in range(detections.shape[2]): # Get the confidence of the prediction: confidence = detections[0, 0, i, 2] # Filter predictions by confidence: if confidence > 0.1: # Get the class label: class_id = int(detections[0, 0, i, 1]) # Get the coordinates of the object location: xLeftBottom = int(detections[0, 0, i, 3] * dim) yLeftBottom = int(detections[0, 0, i, 4] * dim) xRightTop = int(detections[0, 0, i, 5] * dim) yRightTop = int(detections[0, 0, i, 6] * dim) # Factor for scale to original size of frame heightFactor = image.shape[0] / dim widthFactor = image.shape[1] / dim # Scale object detection to frame xLeftBottom = int(widthFactor * xLeftBottom) yLeftBottom = int(heightFactor * yLeftBottom) xRightTop = int(widthFactor * xRightTop) yRightTop = int(heightFactor * yRightTop) # Draw rectangle: cv2.rectangle(image, (xLeftBottom, yLeftBottom), (xRightTop, yRightTop), (0, 255, 0), 2) # Draw label and confidence: if class_id in class_names: label = class_names[class_id] + ": " + str(confidence) labelSize, baseLine = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 1, 2) yLeftBottom = max(yLeftBottom, labelSize[1]) cv2.rectangle(image, (xLeftBottom, yLeftBottom - labelSize[1]), (xLeftBottom + labelSize[0], yLeftBottom + 0), (0, 255, 0), cv2.FILLED) cv2.putText(image, label, (xLeftBottom, yLeftBottom), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 0), 2) # Create the dimensions of the figure and set title: fig = plt.figure(figsize=(14, 8)) plt.suptitle("Object detection using OpenCV DNN module and MobileNet-SSD", fontsize=14, fontweight='bold') fig.patch.set_facecolor('silver') # Show the output image show_img_with_matplotlib(image, "MobileNet-SSD for object detection", 1) # Show the Figure: plt.show()