Research and implementation of improved license plate segmentation and recognition algorithm

License plate recognition system is an important part of information collection in intelligent transportation system, so it is of great practical significance to study it deeply. In this paper, the character segmentation and recognition algorithm of license plate after positioning correction are studied. Aiming at the problems such as the low accuracy of license plate recognition, slow speed of traditional license plate recognition algorithm for complex environment. Aiming at the special license plate or the license plate with stained border, the license plate is divided into seven equal parts by column, and an improved border removal algorithm is proposed, which only deals with the special column of the license plate. In the character segmentation algorithm, the improved projection method is used to realize the accurate segmentation of the disconnected characters. The principle of the algorithm is simple and the processing time is fast. In the character recognition algorithm, the correlation matching method based on Manhattan distance is adopted. The experimental test results show that the algorithm in this paper has improved the accuracy and speed of license plate recognition.