高级检索

  • ISSN 1006-3080
  • CN 31-1691/TQ
引用本文:
Citation:

基于HOG3D描述器与稀疏编码的异常行为检测方法

    通讯作者: 朱煜, zhuyu@ecust.edu.cn
  • 基金项目: 国家自然科学基金(61271349)
    中央高校基本科研业务费专项资金(WH1214015)

  • 中图分类号: TP181

Abnormal Behavior Detection Using Sparse Coding and HOG3D Descriptor

    Corresponding author: ZHU Yu, zhuyu@ecust.edu.cn ;
  • CLC number: TP181

  • 摘要: 提出了一种基于稀疏编码理论的视频异常行为检测方法,并使用HOG3D空-时描述器表征视频序列的形态及运动信息。首先,从正常视频序列中提取空-时兴趣点,获得其特征向量作为训练样本。通过K-SVD字典训练算法构建过完备字典,使得正常样本在所构建字典上的表达具有很好的稀疏性。在稀疏编码过程中,按视频段读取测试视频序列,求解特征信息在字典上的关于其稀疏系数的凸优化问题,然后根据稀疏编码改进公式求得重构误差数值。最后的判断阶段,计算视频段的相对重构误差,相对重构误差为正表明为异常视频段,否则为正常视频段。在UMN数据库3个场景及Weizmann数据库上进行实验,验证了本文算法的有效性。将实验拓展到现实监控视频中,结果表明本文方法在实践中同样具有较好的应用价值。
  • [1] ADAM A,RIVLIN E,SHIMSHONI I,et al.Robust real-time unusual event detection using multiple fixed-location monitors[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(3):555-560.
    [2] KIM J,GRAUMAN K.Observe locally,infer globally:A space-time MRF for detecting abnormal activities with incremental updates[C]//2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2009). Miami,Florida,USA:IEEE,2009:2921-2928.
    [3] KRATZ L,NISHINO K.Anomaly detection in extremely crowded scenes using spatio-temporal motion pattern models[C]//2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2009). Miami,Florida,USA: IEEE,2009:1446-1453.
    [4] MEHRAN R,OYAMA A,SHAH M.Abnormal crowd behavior detection using social force model[C]//2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2009).Miami,Florida,USA: IEEE,2009:935-942.
    [5] LAPTEV I.On space-time interest points[J].International Journal of Computer Vision,2005,64(2/3):107-123.
    [6] DOLLÁR P,RABAUD V,COTTRELL G,et al.Behavior recognition via sparse spatio-temporal features[C]//2nd Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance,2005.USA: IEEE,2005:65-72.
    [7] BLANK M,GORELICK L,SHECHTMAN E,et al.Actions as space-time shapes[C]//Tenth IEEE International Conference on Computer Vision2005,ICCV 2005.Beijing:IEEE,2005,2:1395-1402.
    [8] HELBING D,MOLNAR P.Social force model for pedestrian dynamics[J].Physical Review E,1995,51(5):4282.
    [9] KLASER A,MARSZAŁEK M,SCHMID C.A spatio-temporal descriptor based on 3d-gradients[C]//BMVC 2008-19th British Machine Vision Conference.UK: British Machine Vision Association,2008,275:1-10.
    [10] WILLEMS G,TUYTELAARS T,VAN GOOL L.An efficient dense and scale-invariant spatio-temporal interest point detector[M]//10th European Conference on Computer Vision.Marseille.France:Springer Berlin Heidelberg,2008:650-663.
    [11] CONG Yang,YUAN Junsong,LIU Ji.Sparse reconstruction cost for abnormal event detection[C]//2011 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Providence,RI:IEEE,2011:3449-3456.
    [12] ZHANG Tianzhu,GHANEM B,LIU S,et al.Robust visual tracking via structured multi-task sparse learning[J].International Journal of Computer Vision,2013,101(2):367-383.
    [13] ZHAO B,FEI-FEI L,XING E P.Online detection of unusual events in videos via dynamic sparse coding[C]//2011 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Colorado Springs,USA:IEEE,2011:3313-3320.
    [14] AHARON M,ELAD M,BRUCKSTEIN A.K-SVD:An algorithm for designing overcomplete dictionaries for sparse representation[J].IEEE Transactions on Signal Processing,2006,54(11):4311-4322.
    [15] MEHRAN R,OYAMA A,SHAH M.Abnormal crowd behavior detection using social force model[C]//IEEE Conference on Computer Vision and Pattern Recognition,2009.CVPR 2009.Miami:IEEE,2009:935-942.
    [16] 谢锦生,郭立,赵龙,等.一种基于稀疏编码模型的视频异常发现方法[J].小型微型计算机系统,2014,35(4):917-921.
    [17] 谢锦生,郭立,赵龙.基于快速稀疏编码与惊奇计算的视频异常检测方法[J].中国科学技术大学学报,2013,43(7):561-568.
  • 加载中
计量
  • 文章访问数:  9737
  • HTML全文浏览量:  584
  • PDF下载量:  2346
  • 被引次数: 0
出版历程
  • 收稿日期:  2015-03-24
  • 刊出日期:  2016-02-29

基于HOG3D描述器与稀疏编码的异常行为检测方法

    通讯作者: 朱煜, zhuyu@ecust.edu.cn
  • 1. 华东理工大学信息科学与工程学院, 上海 200237
基金项目:  国家自然科学基金(61271349)中央高校基本科研业务费专项资金(WH1214015)

摘要: 提出了一种基于稀疏编码理论的视频异常行为检测方法,并使用HOG3D空-时描述器表征视频序列的形态及运动信息。首先,从正常视频序列中提取空-时兴趣点,获得其特征向量作为训练样本。通过K-SVD字典训练算法构建过完备字典,使得正常样本在所构建字典上的表达具有很好的稀疏性。在稀疏编码过程中,按视频段读取测试视频序列,求解特征信息在字典上的关于其稀疏系数的凸优化问题,然后根据稀疏编码改进公式求得重构误差数值。最后的判断阶段,计算视频段的相对重构误差,相对重构误差为正表明为异常视频段,否则为正常视频段。在UMN数据库3个场景及Weizmann数据库上进行实验,验证了本文算法的有效性。将实验拓展到现实监控视频中,结果表明本文方法在实践中同样具有较好的应用价值。

English Abstract

参考文献 (17)

目录

    /

    返回文章