The Research Of Moving Vehicles Detection In Video Surveillance

Intelligent transportation system uses the computer vision technology is always a hotissue. It has been widely used in the military, security and other fields. Moving vehiclesdetection and tracking is an important technology of intelligent transportation surveillancesystem. The detection of moving vehicles of video surveillance is refers to judge whetherthere is the motion of the vehicle. It is the foundation of the vehicle recognition andtracking. This paper researches on the moving vehicle detection based on backgroundsubtraction, and puts forward a kind of improved detection algorithm which combiningbackground subtraction with the method of frame difference. The main contents of thispaper include:1) Studies on the current three mainstream moving vehicles detection algorithms,such as optical flow method, frame difference method and background subtraction methodand so on. On the basis of the analysis of these algorithms, this paper summarizes theiradvantages and disadvantages.2) Researches on the commonly used background modeling methods in the movingvehicle detection, and compares and analyses these methods, then obtains their advantagesand disadvantages. And then puts forward the improved background modeling algorithm.3) Based on the background subtraction method, puts forward an impoved movingvehicle detection algorithm combines it with the method of difference frame.4) Writes the simulation source code of the detection algorithm, and through twodifferent complexity videos to get the algorithm simulation and shows the effectiveness ofthe algorithm.5) Operates the detection results using the math morphology techniques in order tocure the target areas.