The Design And Implementation Of License Plate Recognition System Based On Android

With the promotion of smart cities' vision and the expansion of vehicle management requirements,new requirements have been asked for all kinds of vehicle identification systems.But the old license plates recognition system need some specific stationary equipment that can not meet the needs of a flexible intelligent transportation system.such as managing roadside parking and traffic violation.This paper designs and implements a License Plate Recognition APP based on Android platform,which can quickly identify a variety of license plates in complex background without relying on any other third-party libraries.The first step is image preprocessing which include using the phone's stabilization and angle correction function to process image.Second step is loading some preparatory template file.And then using the RGB color threshold to mark some image candidate areas cut by grids.Next,use some morphology knowledge to eliminate the fake plate grid blocks.Through the above steps,the location of the license plates area in the image is confirmed.Then use method based on character outline to segment characters and normalize them.Finally,use the method based on template matching to compare the license character with standard character library data,find the best match as the the recognition result.The last,compile the C++ algorithm source code as dynamic link library to support the recognition function,and android license plate recognition application is completed.