With the increase in the number of vehicles, the license plate recognition system has gradually become a hot spot in scientific research. In this paper, we address the shortcomings of BP neural networks and combine it with the image processing technology of license plate characters, propose a license plate recognition system based on improved BP neural network for license plate recognition. The system mainly includes four modules: license plate image preprocessing, license plate positioning,

more »

... racter segmentation and classification recognition. Firstly, image preprocessing is implemented through image graying, gray stretching, and image binarization. Secondly, edge detection and morphological processing are used to determine the license plate area. Thirdly, the character string of the license plate is segmented. Finally, two corresponding BP neural networks are designed to classify and recognize the license plate characters based on the characteristics of the license plates in China to obtain the recognition results.