The Recognition of License Plate Restrictions Based on Faster R-CNN

Abstract

In this paper, Faster R-CNN was used to recognize the number on the license plate for distinguishing the traffic violation about the license plate restrictions in some cities. Inputting the preprocessed license plate images into a multi-layer convolution neural network, we got the feature maps. Connecting these feature maps to a region proposal network (RPN), the network can output the information about the location and the classification of the license number. With the help of a program that can realize the detection of the last digit of license plate, we can judge whether the car should be restricted or not. The experimental results showed that the Faster R-CNN can recognize the numbers on the license plates in different conditions and judge the last number with good accuracy and robustness.