Research On The Technology Of License Plate Recognition For Intelligent Parking

As the core technology of intelligent traffic management, the license plate recognition has been widely used in the field of road traffic management, intelligent park entrance, parking fees, etc.. License plate recognition improves the efficiency of vehicle management, scientific and intelligent management, and it has a wide range of applications prospects.In this paper, the intelligent parking management system is developed, which focuses on the automatic recognition technology of the license plate, and proposes a super resolution reconstruction algorithm to improve the quality of the license plate.The main research contents include:Firstly, the quality of the image quality of the collected image is the influence of the license plate location and recognition results. This paper analysis and compared a variety of super-resolution image reconstruction method, projection onto convex sets(POCS) super-resolution image reconstruction method, and the theoretical basis of the reconstruction process research. The experimental results show that the image of the license plate is improved, the image quality is higher, and the accuracy of the license plate location is improved.Secondly, this paper analyzes and compares the existing methods of vehicle license plate location. This paper chooses the method of gray scale image plate location based on morphology, which can achieve good results.Thirdly, advantages and disadvantages of the existing character segmentation method are analyzed and compared, this paper chooses a character segmentation algorithm based on projection method and characteristic of the communicating element, The method performs well in the recognition of similar characters.Fourthly, comprehensive analysis and comparison of the existing variety of character recognition methods, proposed a modified template matching method,which first based on the character of rough classification, then the similarity of the grouping template is matched one by one, the method is good for the similar characterrecognition.Finally,using Matlab to carry out the simulation experiments of the key parts of the license plate recognition, license plate location accuracy rate is 94.61%, the accuracy of the character segmentation is 91.71%, the accuracy of Chinese character recognition is 96.45%, the accuracy rate is 95.29%. And the overall test system, the system identification accuracy rate is 90.76%, as in the real environment of the application, the system will further improve the recognition rate. The experimental results show that the method of vehicle license plate recognition is of high accuracy and good performance.