基于SVM的异常行为检测在电子考场中的应用研究  

Research of SVM-Based Abnormal Behavior Detection in Electronic Invigilated Examination Room

作　　者：熊思[1] 

机构地区：[1]湖北第二师范学院计算机学院,武汉430205

出　　处：《湖北第二师范学院学报》2013年第8期42-46,共5页Journal of Hubei University of Education

基　　金：湖北省教育厅科学技术研究计划指导性项目课题的资助(B2013014)

摘　　要：电子监考系统在我国已经得到广泛应用,本文针对考场视频分析的应用需求,以智能视频监控理论为依据,研究基于视频分析的考场异常行为有效识别方法。本文基于目标对象的皮肤特征模型,设计和生成行为模型,结合SVM支持向量机,实现在视频帧序列中检测考场异常行为。该模型可以应用于考场视频监控系统中,有助于提高系统的智能化,增强考场监控能力。Electronic invigilating system has been widely used in China. This paper analyzes the examination room video application requirements, in accordance with the theory of intelligent video surveillance, giving a research of abnormal behavior detection based on video analysis. Based on the characteristics of the target object skin, we design and build a behavioral model, using support vector machine to detect abnormal behavior in the video frame sequence. The model helps to improve the intelligence system, and strengthen the examination room monitoring capabilities.

关 键 词：异常行为 视频分割 动作检测 支持向量机 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]