Design And Implementation Of The License Platerecognition And Traffic Detection System Based On Monitoring System

With the constant improvement of urban infrastructure, intelligent traffic management systems play an increasingly important role in urban traffic management. The wide range of applications of Urban Roads Monitoring System has been promoting the development of intelligent transportation and it became an integral part of traffic management process. This paper is based on the image and video of monitoring system and focuses on the license plate recognition technology and road traffic detection technology. The author developed a prototype system of license plate recognition and traffic detection based on monitoring system with the VC++6.0, Visual Studio and OpenCV. The main work of this paper follows:(1)According to information of monitoring system, the author created a data sample library and divided these information into image information for license plate recognition and video information, mainly for traffic statistics.(2)Using an algorithm that is based on pixel and stable clustering identification to initialize and update the background. The method is based on the stability of the background pixels in the timeline to determine the gray values of pixel. by this method we can get background image with better quality and it facilitates the use of background subtraction.(3)Research on the influence of image preprocessing methods (smoothing, Eliminating noise, binarization, morphological processing, etc.) for image quality and obtain the desired image by different processing.(4) Analysis of the common methods of license plate location, tilt correction, character segmentation and character recognition in License Plate Recognition Process. Improved these methods according to inadequacies of algorithm to improve the recognition accuracy and reduce computation.(5) Analysis of traffic detection algorithm based on the video and improve the algorithm on the basis of the traditional virtual coil method. Using the "centroid substitution" and "shielded signal" to improve detection accuracy, simplify the testing process and improve the real-time detection.(6)Design a prototype system of license plate recognition and traffic detection on urban bayonet system, analysis different modules’ function and design system’s processing, complete software system. To verify the function and stability of the system with the information in the sample library.