Design Of Community Security Management System Based On Video Surveillance

In recent years,with the continuous expansion of the community video surveillance network,the amount of surveillance video data has been growing in spurts,seriously hindering the daily work of video storage,retrieval and analysis.Based on the problems,this paper proposes an intelligent video big data processing method for community surveillance system,which specifically includes:an intelligent storage algorithm for monitoring video driven by abnormal events,an efficient compression algorithm for surveillance video based on depth learning,and a video synopsis algorithm based on retrieval information.At the same time,a community security management system based on video surveillance is designed and implemented by comprehensively utilizing the current mainstream abnormal behavior detection technology and the existing hardware resources of the community.Through the verification of an actual case,the system constructed in this paper can realize credible early warning of abnormal behaviors,efficient compression of surveillance video,selective intelligent storage,and fast retrieval and browsing of video information,greatly improving the in-depth utilization efficiency of surveillance video data.It has highly practical application value.The specific research work of this paper is as follows.（1）For the actual storage requirements of community surveillance video,an intelligent storage algorithm driven by abnormal events is proposed to realize selective intelligent storage of surveillance video.（2）Aiming at the problem that the surveillance video occupies too much memory resources,an efficient compression algorithm for surveillance video based on depth learning is proposed,which can both improve the video compression ratio and ensure the decompressed video quality as much as possible.（3）For the problem of slow manual browsing and retrieval of massive surveillance videos,a video summarization algorithm based on retrieval information is proposed,which generates corresponding video summaries according to users’ retrieval information.It can save users’ retrieval time and enable users to quickly and efficiently screen out information of interest from massive surveillance data.（4）Taking a real residential project as an example,on the basis of the demand analysis of the community video surveillance security management system,the hardware architecture and software architecture are designed and built,and the performance of the system is verified by tests and experiments,thus indicating the practical application value of this system.