基于自适应高斯模型和运动能量的异常行为识别  

Abnormal Behavior Recognition Based on Adaptive Gaussian Model and Motion Energy

作　　者：赵雪章[1] 吴嘉怡 席运江[2] ZHAO Xuezhang;WU Jiayi;XI Yunjiang(Electronic Information School,Foshan Polytechnic,Foshan 528137,China;School of Business Administration,South China University of Technology,Guangzhou 510641,China)

机构地区：[1]佛山职业技术学院电子信息学院,广东佛山528137 [2]华南理工大学工商管理学院,广东广州510641

出　　处：《现代信息科技》2023年第19期79-82,88,共5页Modern Information Technology

基　　金：国家自然科学基金项目(72171090);广东省教育厅创新类项目(2019GKTSCX119);广东省教育厅教育教学改革研究与实践项目(GDJG2019023)。

摘　　要：为有效识别监控视频中的群体异常行为,提出一种基于自适应高斯模型和运动能量的异常行为识别方法。将自适应帧间差分法融入混合高斯模型中,对运动目标进行提取,计算行为发生个体的动态能量,利用行为发生各方的位置关系计算出交互能量,最终计算出异常行为事件的整体能量总值,从而实现群体异常行为的有效识别。实验结果表明,该文算法对人群异常行为具有较好的识别效果,算法实时性较好,具有一定的应用推广价值。To effectively identify group abnormal behavior in surveillance videos,a method for identifying abnormal behavior based on adaptive Gaussian model and motion energy is proposed.Integrating the adaptive inter frame difference method into a mixed Gaussian model,extracting moving targets,calculating the dynamic energy of individuals involved in behavior,utilizing the positional relationships of all parties involved in behavior to calculate interaction energy,and ultimately calculating the overall energy total of abnormal behavior events,thus achieving effective identification of group abnormal behavior.The experimental results show that the algorithm proposed in this paper has good recognition performance for group abnormal behavior,and has good real-time performance,which has certain application and promotion value.

关 键 词：高斯模型 运动能量 异常行为 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] TP312[自动化与计算机技术—计算机科学与技术]