Research And Implementation On Motion Detection In Network Real-Time Audio And Video System

With the development of computer network, digital image processing and transmitting technology, network audio-visual monitoring applications are increasing rapidly such as video conferencing system, VOD, distance education, remote medical treatment, remote monitoring and battlefield directing. It is predicted that network audio-visual monitor would be more and more important on both production and daily life.Network real time audio and video system is a complex system, which involves digital image processing, computer network, hardware and software. This paper mainly focuses on the technology of real time audio and video in the background of computer network.Motion detecting is a basic function in video monitoring system. The paper studies how to do motion detecting with high efficiency. By using mathematical model of digital image processing, this paper optimizes a traditional algorithm and implements that optimized algorithm. Moving target detecting and tracking is another basic function in video monitor system, which is in the lowest level of vision processing and will be the foundation of all high level processing. This paper adopts auto-adaptive background subtraction arithmetic on single static video node and detects moving areas under the complicated and dynamic background. Based on the analysis of the static and dynamic characters of the moving target, this paper improves the correctness of moving target tracking. Later on, this paper discusses and presents a model of tracking system under single movable node.According to the characters of the network real-time audio and video system, this paper presents the architecture of the system and elaborates functionalities of data acquisition subsystem, MPEG4 compression subsystem and streaming media transmission subsystem.