Python Image Processing Cookbook

Getting started

First, download a pretrained model for the Faster R-CNN (for example, faster_rcnn_resnet101_coco, with ResNet101 as the backbone and trained on the MS-COCO dataset) from the tensorflow object detection model zoo (https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md) and extract/save .pb of the model (frozen inference graph) inside the appropriate location in the models folder. Let's now import all of the required packages using the following code snippet:

import numpy as npimport cv2from PIL import Image, ImageFont, ImageDrawimport json import colorsysimport matplotlib.pylab as pltimport tensorflow as tfprint(tf.__version__) # 2.0.0