The Research Of Human Video Tracking Based On MeanShift Algorithm

The detection and tracking of human motions is one of the most active topics in computer vision. As U.N. attaching the importance to the social public safety, the safe intelligent detection system needs to be improved and become more and more effective in public. More and more human and financial resources from many well-known companies and research institutions pay for the research of the intelligent monitoring system. This topic can be predicted to have broad prospects and to bring about tremendous social benefits in future.In this thesis, a method of human motion detecting and tracking based on video is proposed. These three steps are inseparable; the first step is a prerequisite to achieve after the step when the programming is in accordance with the order described in this article step by step chapter completed. This major work has done now summarized as follows:1. A detailed description of the visual analysis of human movement background, status and application of practical significance, discussed the advantages and disadvantages of various methods of human body tracking and human tracking in difficult problems.2. The moving target detection and identification of the human body, the details of the three types of target detection and compare it. Analyzed based on a fixed single camera under the conditions of how to improve the background model, using a dynamic threshold segmentation of foreground regions, so that mutations in the case of light can detect movement very good prospects.3. In the tracking process of moving target, deeply into the study and research of the non-participation theory and MeanShift density estimation theory, a projection will focus on integrating and space tracking algorithm is discussed and explored for the purpose to improve the tracking stability improved methods of tracking human body.4. A system of human detection and tracking based on Visual C++6.0 is designed. It has a friendly interface. Multifunctional modules are involved. The involved knowledge of image processing, pattern recognition can be realized.