Research On Moving Object Classification In Intelligent Video Surveillance

Along with the development of the society, economy and the computer technology,people pay more attention to the security problems in their lives. So the video surveillance hasbeen greatly popular now. As a result of the popularity of surveillance, a large amount ofvideo data was generated. In the past, we could hire special security guard to monitor thesurveillance. But it becomes impossible because there is too many data too handle for human.So the surveillance video data is always used for investigation after accidents rather thanmonitoring to prevent accidents. More unfortunately, massive video data for laterinvestigation is also a very heavy and boring job and it needs many people to spend plenty oftime to deal with this screening job. Therefore, people pay great attention of the intelligentvideo surveillance technology. The demand of making computer instead of the human tomonitor the surveillance video data becomes more and more big. In intelligent videosurveillance technology, the technology of moving target classification is particularlyimportant. The ultimate goal of intelligent video surveillance system is to understand thebehavior of specific target in the video, to discover the potential unsafe factors, alert to theoutside world. However, there are a lot of moving objects in surveillance video, such aspedestrians, vehicles, leaf, noise, etc., to understand the behavior of a moving target that weare interested in, the first thing to do is to determine the type of the movement of the target.This paper mainly aims at the classification of the moving target in long distance shootingsurveillance video. In this case, the moving target is small and fuzzy. In this paper, it willintroduce some technology about the classification of moving target in intelligent surveillancefirst, then a classification algorithm will be put forward using the technology that have justsaid, finally an experiment result will be presented to show how the algorithm works andsome analysis of the experiment’s result.