The Research Of Object Detection And Tracking Based On ITS Vehicle Video

In recent years, the traffic flow in some big cities becomes bigger and bigger with the acceleration of urbanization. While enjoying the great convenience of cars, we face a lot of traffic problems. Under certain circumstances, it is a very effective way to establish comprehensive intelligent transport system. Motion object detection and tracking is the main research field of computer vision and image coding, which has been widely applied in military, medicine and scientific research etc. The accuracy and stability of tracking effect depend on the design of algorithms to a great extent. It is still a hot point to research an accurate and robust algorithm in computer vision.This paper establishes an object monitoring and tracking system based on intelligent transport system. Its main function is to monitor moving objects from fixed settings and track the object, which is automatically chosen. This paper uses theories such as background subtraction, Kalman filter, mean shift method and so on. These methods further improve the intelligence, real-time ability and adaptability.The major work implemented in this paper is presented as follows:1. The monitoring and segmentation of sequence images: This paper compares the differences between optional flow, differential between adjacent frames and background subtraction method. It also conducts many kinds of filtering process such as average, mean and mathematical morphology filter to images and brings up a relatively effective method to segment and subtract moving objects.2. The tracking of sequence images: According to the former segmentation results, this paper makes a comparison between two kinds of tracking methods, Kalman and mean shift methods. It finally realizes the automatic tracking of moving objects in a fixed setting. The experiment result shows the feasibility of this method.