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

With the development and progress of society,more and more road vehicles,the society put forward higher demands on intelligent transportation.The degree of intelligence of road traffic depends on the development and progress of video image processing and machine learning technology.Vehicle tracking and license plate recognition technology is a very important key technology in vehicle identification system.Through vehicle tracking technology,it can obtain the position information of sports vehicles,Such as data and information,in order to achieve the positioning of vehicles.And license plate recognition technology can extract the vehicle’s unique logo license plate number.So the vehicle tracking and license plate recognition technology based on embedded devices designed and implemented,which has practical significance and practical value to promote intelligent traffic and vehicle identification system of technological progress.This paper presents a license plate recognition system that is based on Android and I.MX6 Q embedded processor.The whole system is divided into two parts: vehicle tracking and license plate recognition.The mobile phone client is connected with the vehicle tracking system,and the vehicle tracking system can track the moving vehicle.When the moving vehicle reaches the region of interest,the vehicle image is intercepted and sent wirelessly to the mobile client.When the mobile client receives the vehicle image,the license plate recognition algorithm is used to realize the extraction of the license plate number in the vehicle image.The vehicle tracking system is based on NXP Semiconductors I.MX6 Q quad-core processor,Android4.3 system as the platform and Harbin Engineering University Cultural Park as the realistic background.The CMOS image sensor OV3640 is used to collect the video images of the moving vehicle and the recorded motion.The vehicle video is stored in the built-in storage for system reading and analysis.The vehicle tracking algorithm is based on PBAS motion target detection algorithm and blob matching method.The license plate recognition system is based on the realization of the mobile client.The license plate recognition is divided into two parts: the license plate location and the character recognition.The license plate positioning using the sobel operator to extract the edge feature.The SVM support vector machine is used for template training.The character recognition part adopts the histogram of the image as a feature,and adopts multi-layer neural network to template training.This paper introduces the background,significance and present situation of the license plate recognition system,and then introduces the design scheme and the related technology,the design and implementation of the vehicle tracking system,the design and implementation of the license plate recognition system.Finally,the system is tested and verified.The experimental results show that the design achieves the desired target.