Research And Implementation Of Automatic License Plate Recognition System In Complex Environment

With the maturity of big data technology and the vigorous development of artificial intelligence technology,smart transportation has gradually become an important place for related technologies.Through intelligent transportation technology,the road network can achieve higher operational efficiency,which saves people's time and reduces resource consumption.The application of the urban surveillance system enables the public security organs to track vehicles and pedestrians more quickly,thus making our lives more secure.The license plate is the unique identifier of the vehicle,so the automatic identification technology of the license plate is a supporting technology for smart transportation and plays a very important role.This paper describes an automatic license plate recognition system consisting of a license plate location module,a character segmentation module,a character recognition module and a split-free license plate recognition module.This paper first analyzes the background information and research significance of the license plate automatic identification system.The second chapter introduces the related theory and the development of automatic license plate recognition technology.The third chapter introduces the license plate recognition algorithm scheme adopted in this paper,and compares the method adopted in this paper with other mainstream methods.The fourth chapter describes the demand analysis,summary design and detailed design of the license plate automatic identification system.The system achieves the expected performance under the unified data set.Finally,in the fifth chapter,the work of this paper is summarized and the follow-up work is prospected.The main work and innovations of this paper are summarized as follows:1.A comprehensive and in-depth investigation and summary of the development status of automatic license plate recognition technology,which is divided into four stages:license plate location,character segmentation,character recognition,and segment-free license plate recognition.The mainstream methods at each stage and their advantages and disadvantages were clarified,and experimental comparisons were made using a unified data set.2.The traditional Haar+Adaboost license plate location algorithm is improved,and a CNN binary classifier is introduced to further remove the license plate image,forming a Haar+Adaboost+CNN license plate location scheme.3.Improve the character segmentation scheme,using a sliding window method to segment characters.4.Designed and implemented a simple automatic license plate recognition system,and tested the system with a unified data set,the system achieved the expected performance under the test set.