一种基于速度强度熵与纹理特征的人群异常检测算法  

An anomaly detection algorithm based on VMME and texture features

作　　者：李斐[1] 陈恳[1] 李萌[1] 郭春梅[1] LI Fei;CHEN Ken;LI Meng;GUO Chun-mei(Faculty of Electrical Engineering and Computer Science, Ningbo University, Ningbo 315211, China)

机构地区：[1]宁波大学信息科学与工程学院,浙江宁波315211

出　　处：《宁波大学学报（理工版）》2017年第4期63-67,共5页Journal of Ningbo University：Natural Science and Engineering Edition

基　　金：国家自然科学基金(60972063);宁波市自然科学基金(2014A610065);宁波大学学科项目基金(XKXL1308)

摘　　要：人群异常事件检测是智能视频监控领域的重要研究内容,文章提出了一种融合速度强度熵VMME与纹理特征的人群异常行为检测算法.该算法采用LBPCM算法提取图像纹理特征,在视频帧计算光流基础上,获得特征点速度强度图,并以其熵VMME作为场景运动特征,将场景纹理特征和运动特征送入支持向量机训练分类.实验表明,新算法可实现对人群异常行为的检测,且有较高准确率.In the field of intelligent video surveillance,the detection of abnormal events has remained animportant subject in the research field.This paper proposes an algorithm for detecting abnormal behavior ofcrowds based on entropy of velocity magnitude map(VMME)and texture feature.Through the computation ofoptical flow in a video frame,the velocity magnitude map of feature points can be obtained,and the entropy ofvelocity magnitude map can be calculated as the feature of scene motion.LBPCM is first used to extract thetexture features of the crowd,then the features of two kinds are fused into the support vector machine fortraining classification.Experiments show that the algorithm can effectively detect abnormal behavior and hashigh detection accuracy rate.

关 键 词：人群异常检测 纹理特征 运动特征 VMME LBPCM 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]