Detection Of The Moving Object And Face Tracking Based On Serial Images

Intelligent video surveillance technology is one of today's cutting-edge research topics. Based on the foundation of image processing and computer vision, it is able to detect, track and identify the moving targets in monitoring picture or situation, even to analyze the behavior of target objects. Therefore, it plays an important role in safety precautions, traffic management and incident response. And the technology for moving target detection and identification, as the core part of intelligent video surveillance technology, is attracting more and more attention, which has become the focus of study. This paper mainly studied on moving target detection and face tracking technology, the key of an intelligent video surveillance system.On moving target detection, the paper studied the detection and segmentation on the moving target in a simple scene, and analyzed the advantages and disadvantages of several typical algorithms, including optical flow method, inter-frame difference method and background difference method. And an algorithm for moving target detection in a fixed background was built accordingly, which is developed from the inter-frame difference method and completes moving targets detection by using multi-Gaussian background modeling.On face tracking, the paper studied the basics of face detection and tracking, analyzed the AdaBoost-based face detection algorithm and MeanShift-based face tracking algorithm in particular, which implemented the straightforward frontal face detection and tracking in the serial images..It has been found that the combination algorithm of inter-frame difference method and multi-Gaussian background modeling method proposed in the paper could detect the moving target fast, accurately and completely, and achieve good results in tracking of frontal detection of moving face.