Surveillance Video Target Tracking Algorithms And Implement Of Application System

With the development of economy, society, and the improvement of technology, the video surveillance industry rapidly develops. In recent years, video processing theory and technology has been a hot research. Due to the performance boost of related embedded hardware, make these studies more practical value and social significance.This article includes two parts, one study single moving target tracking in surveillance video process, researching Mean Shift algorithm and particle filtering algorithm in target tracking applications, and then analysis the advantages and disadvantages of these algorithms and related parameters; the second part is about developing a dome system based on a Davinci architecture dual-core processor, the DM6446. In this platform we can fully take advantage of the DSP core as well as the Codec Engine framework for video processing, the ARM core is based on the Linux operating system, used to develop applications. In addition, we build a Web server in this embedded system.