公共治安监控网络图像高敏感信息预警仿真  

Early Warning Simulation of Image Highly Sensitive Information in Public Security Monitoring Network

作　　者：王燕 刘琳芳 WANG Yan;LIU Lin-fang(Investigation Department of Guizhou Police College,Guizhou Guiyang 550005,China;School of Mathematics and Statistics,Guizhou University,Guizhou Guiyang 550005,China)

机构地区：[1]贵州警察学院侦查系,贵州贵阳550005 [2]贵州大学数学与统计学院,贵州贵阳550005

出　　处：《计算机仿真》2020年第6期420-424,共5页Computer Simulation

基　　金：贵州省教育厅青年科技人才成长项目(黔教合KY字[2018]284)。

摘　　要：针对视频监控系统敏感信息检测存在局限性大的问题,提出一种公共区域治安监控网络高敏感信息预警方法。首先构建混合高斯监控背景模型,并依据场景的转换,实时模型更新;通过光流法中梯度检测法对高敏感个体快速检测和定位,引入H-S算法与L-K算法,分析高敏感信息状态变量,并获取给定区域的概率密度,以概率密度结果为依据,实现高敏感信息分布情况的计算,即高敏感信息预警。通过仿真得出,所提方法能够及时有效的对存在高敏感信息的区域预警,且检测错误率、误报率低,整体方法运算简单有很强的自适应能力。Due to high limitations in sensitive information detection,this article focuses on an early warning method for high-sensitive information in public security monitoring network.Firstly,the mixed Gaussian monitoring background model was constructed,and the real-time model was updated by the scene transformation.Through the gradient detection method in optical flow method,the high-sensitive individuals were detected and located quickly.Moreover,H-S algorithm and L-K algorithm were used to analyze the state variables of high-sensitive information,and then the probability density of the given region was obtained.Based on the probability density,the distribution of high-sensitive information was calculated.In other words,the early warning of high-sensitive information was realized.Simulation results show that the proposed method can timely and effectively achieve the early warning of high-sensitive information.The detection error rate and false alarm rate are low.The overall method is simple.Meanwhile,this method has strong adaptive ability.

关 键 词：公共区域治安监控 网络高敏感信息 光流法 预警 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]