Traffic Violation Monitoring System Designed Based On Video Sequence

Intelligent transportation system(referred to as ITS) is the future development direction of transportation system, which effectively integrates advanced information technology、control technology and computer technology, applying them to the entire ground traffic management system. The important parts of ITS is traffic violation snapshot and vehicle tracking and traffic statistics system. The traditional traffic system based on ground induction coil is not only a large project to install and difficulty to maintain, but also its function is limited, what’s more, it can not meet the needs of modern traffic. Therefore, it is importance to design an integrated system based on video sequence, which includes vehicle tracking, illegal capture and traffic statistics function.In this paper, it is firstly to analyze the principle involved in the Traffic violation monitoring system based on video sequence, and then study the theory and algorithms contained image preprocessing, target detection and tracking, eliminate shadows and traffic statistics. Through simulating and analysis each step’s algorithm, comprehensive simulation results and system operation speed. Eventually adopting median filter for pre-processing video sequence, with Pyramid optical flow(L-K) algorithm for target detection, and HSV color space to eliminate the shadow of moving objects, the target tracking based spatio-temporal context information(STC). The processing core of system’s hardware platform is built based on ARM11 Raspberry Pi. Micron’s MT9P031 CMOS sensor is used for image acquisition design. Expansion module is designed for the system providing power and protection. Finally, Use Qt and Open CV which have python interface to realize the integrated vehicle tracking, illegal capture and traffic statistics, illegal snapshot software. Through field test results show that: the system is reliable, stable, and has practical value.