Research On Video Vehicles Detection Algorithm In ITS

These years, the great economic benefit has been gained through ITS.And ITS becomes more and more important to strengthen civil traffic mangment. The technology of video traffic information detection discussed in this paper plays an important part in the ITS.Through image analysis and processing it can obtain much useful traffic information that is key part of the automatization traffic management.In the paper a method based on virtual detection line is used to detect the vehicles.Firstly, the system based on DSP is introduced. Then the camera calibration algorithm is detailed. It achieves the convert of from image coordinates to world coordinates.It is the base of vichile detection.Then the vehicle detection algorithm is discussed in detail. Vehicles can be detected with the virtual line. Through the method many traffic information can be achieved. The information consisit of vehicle detection,velocity detection,vehicle number detection,vehicle type detection,transport density,road occupancy,and so on. And then algorithm flow and program design is described.At last, the algorithm to fuse visible image and infrared image is introduced and studied. Through the technology the accuracy of vehicle detection can be improved.It can make up for the shortage of traditional vedio detection and will have a good development.