欢迎访问中国科学院大学学报,今天是

中国科学院大学学报 ›› 2013, Vol. 30 ›› Issue (1): 83-89.DOI: 10.7523/j.issn.1002-1175.2013.01.013

• 计算机科学 • 上一篇    下一篇

基于时空惊奇计算的视频异常检测方法

谢锦生, 郭立, 陈运必, 赵龙   

  1. 中国科学技术大学电子科学与技术系, 合肥 230027
  • 收稿日期:2011-10-10 修回日期:2012-03-21 发布日期:2013-01-15
  • 通讯作者: 郭立
  • 基金资助:

    国家自然科学基金(61071173)资助

Anomaly detection method in video based on spatio-temporal surprise computation

XIE Jin-Sheng, GUO Li, CHEN Yun-Bi, ZHAO Long   

  1. Department of Electronic Science and Technology, University of Science and Technology of China, 230027, China
  • Received:2011-10-10 Revised:2012-03-21 Published:2013-01-15

摘要:

提出一种基于贝叶斯惊奇计算的视频异常检测方法.用块匹配运动估计方法提取运动特征(如运动幅度、方向),得到多尺度运动矢量直方图.使用空间维度与时间维度上惊奇计算相结合的度量方法,既可以检测"个体异常行为",也可以应用于"群体异常行为"检测.实验表明,该算法是鲁棒和实用的,且易于实现.

关键词: 视觉注意模型, 视频分析, 贝叶斯惊奇理论, 异常检测

Abstract:

We propose an anomaly detection method in video based on Bayesian surprise computation. We use the block-matching motion estimation method to extract low-level motion features (such as magnitude and direction of motion) and then calculate multi-scale histogram of motion vector. We use both spatial surprise and temporal surprise to detect not only "individual abnormal behavior" but also "group abnormal behavior". Experimental results show that our algorithm is robust and applicable and it can be easily implemented.

Key words: visual attention model, video analysis, Bayesian theory of surprise, anomaly detection

中图分类号: