Design And Implementation Of A Driving-test System Based On Video Surveillance Technology

Video surveillance technology is not only a multi-disciplinary scientific subject, but also a technology which is currently widely used. Video surveillance is a hot topic nowadays, with the development of computer technology and image processing technology, video surveillance technology more and more infiltrates into entertainment, education, health, security, sports and other fields. In general, video surveillance technology presents a trend of digitalization, networked, integrated, and intelligent.In this paper, through the analysis and research of traditional driving-test system, combined with video surveillance technology, we worked hard on angle point method and particle swarm optimization theory, and achieved the object that automatically tracking and monitoring moving vehicle. Based on the theory, we designed a complete set of driving-test system which would work accurately and correctly under any weather conditions.This paper includes following contents:On the basis of the analysis of the demand and current situation of the traditional driving-test system, firstly, we proposed a vehicle contour extraction algorithm based on corner point method, analyzed the pros and cons of linear projection method and SUSAN algorithm, and improved the SUSAN algorithm in adaptive threshold; secondly, we described three kinds of algorithms used in video surveillance system, as known as optical flow field method, detection method of background subtraction, and detection method of Interframe differentiation, and according to the evaluation of the advantages and disadvantages of each algorithm, we presented a vehicle tracking algorithm based on particle swarm optimization; finally, on the basis of all above, we designed a whole set of software algorithm and the driving-test system based on video surveillance technology using Microsoft Visual Studio 2008 development tools in Windows environment.According to the experimental results, the improved SUSAN detection algorithm and the particle swarm optimization algorithm can well meet the requirements of vehicle recognition and tracking, the recognition ability of our driving-test system based on these algorithm has the advantage of high stability, fast speed, and good effect. The high accuracy driving-test system is adapted to any vehicle in any weather conditions, and improved the fairness, justice, accuracy of traditional driving-test systems. It completely reached our expected goal.