Design And Implementation Of Intelligent Video Abnormal Behavior Surveillance System

In recent years, with the development of video surveillance, the detection and analysis ofabnormal behavior is an increasingly important research and application topic. On basis ofanalyzing the intelligent video surveillance technology and abnormal behavior algorithms, thispaper propose a design and implementation of intelligent video abnormal behavior surveillancesystem. This article focuses on the following aspects:(1) In the objects extraction, this paper gives acombination method of the inter-frame difference algorithm and the background reductionalgorithm which used for extracting the moving target. Then, based on FNCC algorithm is adoptedto eliminate the shadow. At the same time, by using the Multi-feature matching algorithm based onKalman filter it can realize the target tracking.(2) In the analysis of abnormal behavior, this paperputs forward the recognition algorithms concerning the Geometric characteristics to detect fourkinds of common abnormal behavior which includes unknown objects left behind, moved objects,crossing the line and suspicious wander.(3) The consist of system as well as the function andimplementation processes of each components, are detailedly introduced in this paper. The mainfunction of the system are as follow: obtain and play the video streaming, display the video-relatedinformation, select the abnormal behavior recognition algorithm, set the detection area as well asthe rules, push the alarm information in real-time, inquiry alarm information intelligently andplayback the alarm information video by using key-frame.The experimental results of the system show that it can be used in place such as: residentialquarters, squares, school and so on to recognize the abnormal behavior effectively whichdemonstrates that this system has mainly achieved the expected goals.