Python Image Processing Cookbook

How to do it...

Perform the following steps to implement object detection using the pretrained Faster R-CNN network in TensorFlow:

1. Initialize the list of class labels that the MobileNet SSD model was trained to detect:

with open('models/image_info_test2017.json','r') as R: js = json.loads(R.read())labels = {i['id']:i['name'] for i in js['categories']}print(len(labels))# 80

2. Generate a set of unique colors for the bounding boxes for each class:

HSV_tuples = [(x/len(labels), 0.8, 0.8) for x in range(len(labels))]colors = list(map(lambda x: colorsys.hsv_to_rgb(*x), HSV_tuples))

3. Read the input image and the pretrained TensorFlow model's frozen inference graph:

img = cv2.imread('images/bus.jpg')with tf.io.gfile.GFile('models/faster_rcnn/frozen_inference_graph.pb', ...