|
As an important application form of target automatic recognition, vehicle license plate recognition (LPR) technology can be used in many fields, such as traffic monitoring, accident investigation, traffic violations record, automatic fare collection, access control and many other occasions, this can improve the degree of automation of traffic management, more and more attention related to this technology has been paid.License plate location technology is one of the core technologies in the license plate recognition system, this paper introduces three license plate location algorithms. The first method is the traditional algorithm using jump variables. The second is a license plate location algorithm under color space, the positioning method includes calculating the value of the color image features, image binarization, morphological operations and the extraction of the license plate. The algorithm can locate license plate accurately under different light conditions in complex background. The third method is using Adaboost algorithm to locate license plate. Then we introduced a new character segmentation method, including the character segmentation in both horizontal and vertical direction, character normalization, smoothing and other algorithms. Finally, using BP neural network for character recognitionThe system is achieved in the C++ environment, and it uses images under different lighting conditions to test, the code is packaged into a DLL library and applied in electronic police. Test results show that this method is feasible, fast, accurate and robust, to meet the requirements of license plate recognition system, has some practical value.
|