Moving Object Detection And Tracking In Outside Environment

Moving object detection is the premise of computer vision. Utilizing the technology of image processing and pattern recognition, by image processing of video sequences, it detects and locates moving object, moreover, it may identify, track, and further have an automatic analysis to behavior properties of the target, then issued early warning of suspicious behavior. As a key basis work for technology of intelligent video surveillance, moving object detection has been obtained quite extensive and in-depth research. But the research largely stays in the background of a relatively simple fixed scene at home and abroad. It is not enough on the research of the complex dynamic background, while now due to people’s lives environment is complicated and changeable, so the major challenges which the video surveillance system faces is how to accurately and efficiently detect moving objects in the complex background.The main content of this paper is the research of the moving object detection and tracking algorithm in outside environment. The article first introduces the classic moving object detection algorithm and major difficulties moving object detection faces in outside environment, then an improved Gaussian mixture background modeling algorithm is put forward. Aiming at the insufficient of Gaussian mixture model, the improved algorithm divides the objects in the scene into three categories: background, middle-ground and foreground. Moving objects including foreground and middle-ground are extracted firstly; then foreground is segmented from middle-ground. In this way, almost middle-ground is filtered, so we can obtain a clear foreground objects. Thus, it overcomes the interference of the dynamic background in outside environment. The improved algorithm makes the test results accurate and clean. Experimental results show that the proposed algorithm can detect moving objects much more precisely, and it is robust to lighting changes and shadows.The end of this article also describes the status of the moving object tracking algorithm. After the moving object detection, due to it may fail in next detection, it is necessary to track and locate. This paper first introduces the basic principles of the moving object tracking algorithm as well as some mainstream tracking algorithm. And it focuses on the research of moving object tracking algorithm based on Kalman filtering. Finally it implements the moving property of the moving object tracking algorithm based on Kalman filtering.