Background Subtraction And Shadow Detection Method And Achieve. Video Traffic

Intelligent Transportation System (ITS) is the direction of development in the future, which services extensively in the transport system, and provides real-time, accurate and efficient traffic information to bring the great convenience for people's manufacturing and life.First, the Intelligent Transportation System captures the traffic information through real-time visual detection, and then deal with the videos to control and manage the traffic.This article bases on real-time collection of traffic video information and takes the vehicles in video for the main object of study. Though the extraction and update of video background, it achieves the detection of the moving vehicles'shadow. Eventually it develops a prototype video vehicle detection to achieve the detection of moving vehicles.In terms of extraction and updating the background, after comparing and studying to the existing methods of extracting background, the traffic flow and speed limits will be affected by using the method of the average to extract the background.This article provides an improved method of extracting the background, which is updating the background of the algorithm in using of the difference between the averages algorithms. This method is faster and more efficiently in extracting the background image, and provides a strong protection for the follow-up work.In terms of object detection, the image difference method is mainly used to get the current moving, and detect moving object and the shadow region for better detecting shadows.For the shadow extraction, through comparing and studying the principle of HSV shadow detection, a matrix of HSV-based shadow detection is provided in this article. The speed and accuracy of detecting target shadow will be improved effectively by the HSV detection in a number of small regions through matrix block.The test shows that the method of the scene fully meets the system requirements with high practicality in the fast way.The proposed method can reduce the computation process in the video detection and meet the requirements of real-time in the quick way. Experimental results showed that the proposed algorithm have a great performance extracting the background, shadow vehicle inspection and testing the accuracy of the robust.During the design and implementation, Visual C++and Open CV are used in developing the digital video processing system, and it realized the background video image extraction, shadow detection and many other features.