Video Intelligent Monitoring Based on Object Segmentation and Behavior Warning
HAO Ping
Abstract：In order to solve the problem that the current video surveillance system is complex and changeable,the video intelligent analysis is wrong,the video surveillance and intelligent analysis real-time system based on adjacent frame difference and background difference is proposed in this paper.Firstly,based on the analysis of the principle of adjacent frame difference algorithm and background subtraction algorithm,the advantages of the two algorithms are analyzed,and the video object segmentation algorithm is established.Then,according to the characteristic of the abnormal behavior event video image,set up the abnormal characteristic threshold,set up the alarm monitoring mechanism.Finally,based on the C# language and the Aforge.NET open source library,the development and implementation of this algorithm and system.Experimental results show that:compared with the current video surveillance system,in the face of the target imaging law is not stable,the algorithm has a higher monitoring accuracy.
Keywords：adjacent frame differencebackground subtractionvideo surveillanceimage featureAforge.NET
Publication Date：2017-01-01
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：5( 1327-1331 )