基于时空兴趣点和概率潜动态条件随机场模型的 在线行为识别方法

doi:10.11772/j.issn.1001-9081.2017112805

计算机应用 ›› 2018, Vol. 38 ›› Issue (6): 1760-1764.DOI: 10.11772/j.issn.1001-9081.2017112805

• 虚拟现实与多媒体计算 • 上一篇下一篇

基于时空兴趣点和概率潜动态条件随机场模型的在线行为识别方法

吴亮, 何毅, 梅雪, 刘欢

南京工业大学电气工程与控制科学学院, 南京 211816

收稿日期:2017-11-29 修回日期:2018-01-05 出版日期:2018-06-10 发布日期:2018-06-13
通讯作者: 何毅
作者简介:吴亮(1992-),江苏宿迁人,硕士研究生,主要研究方向:视频序列、图像处理与识别;何毅(1969-),男,江苏扬州人,副教授,博士,主要研究方向:嵌入式系统;梅雪(1975-),女,内蒙古呼伦贝尔人,副教授,博士,主要研究方向:计算机视觉、图像分析与理解、模式识别、机器学习;刘欢(1990-),男,江苏徐州人,硕士研究生,主要研究方向:图像处理。
基金资助:
江苏省"六大人才高峰"项目（XXRJ-012）；江苏省研究生科研与实践创新计划项目（SJCX17_0276）。

Online behavior recognition using space-time interest points and probabilistic latent-dynamic conditional random field model

WU Liang, HE Yi, MEI Xue, LIU Huan

College of Electrical Engineering and Control Science, Nanjing Tech University, Nanjing Jiangsu 211816

Received:2017-11-29 Revised:2018-01-05 Online:2018-06-10 Published:2018-06-13
Supported by:
This work is partially supported by the Six Talent Peaks Project in Jiangsu Province (XXRJ-012), the Postgraduate Research and Practice Innovation Project in Jiangsu Province (SJCX17_0276).

摘要/Abstract

摘要： 针对在线行为连续序列的识别问题以及行为识别模型的稳定性问题，提出一种监控视频中基于概率潜动态条件随机场（PLDCRF）的在线行为识别方法。首先，应用时空兴趣点（STIP）对行为特征进行提取；再利用PLDCRF模型识别室内人体的活动状态。PLDCRF模型融合了隐含状态变量，能够构建姿态序列子结构，可以选取姿态之间的动态特征，并且直接标记出未分割序列；同时也可以正确地标记出行为间的转换过程，从而明显改善了行为识别的效果。隐含条件随机场（HCRF）、潜动态条件随机场（LDCRF）、潜动态条件神经场（LDCNF）以及PLDCRF模型对10种不同动作的识别率比较结果表明，所提PLDCRF模型对连续的行为序列的综合识别能力更强，并且有更好的稳定性。

关键词: 视频监控, 在线行为识别, 时空兴趣点, 概率潜动态条件随机场

Abstract: In order to improve the recognition ability for online behavior continuous sequences and enhance the stability of behavior recognition model, a novel online behavior recognition method based on Probabilistic Latent-Dynamic Conditional Random Field (PLDCRF) from surveillance video was proposed. Firstly, the Space-Time Interest Point (STIP) was used to extract behavior features. Then, the PLDCRF model was applied to identify the activity state of indoor human body. The proposed PLDCRF model incorporates the hidden state variables and can construct the substructure of gesture sequences. It can select the dynamic features of gesture and mark the unsegmented sequences directly. At the same time, it can also mark the conversion process between behaviors correctly to improve the effect of behavior recognition greatly. Compared with Hidden Conditional Random Field (HCRF), Latent-Dynamic Conditional Random Field (LDCRF) and Latent-Dynamic Conditional Neural Field (LDCNF), the recognition rate comparison results of 10 different behaviors show that, the proposed PLDCRF model has a stronger recognition ability for continuous behavior sequences and better stability.

Key words: video surveillance, online behavior recognition, Space-Time Interest Point (STIP), Probabilistic Latent-Dynamic Conditional Random Field (PLDCRF)

中图分类号:

TP391.4

吴亮, 何毅, 梅雪, 刘欢. 基于时空兴趣点和概率潜动态条件随机场模型的在线行为识别方法[J]. 计算机应用, 2018, 38(6): 1760-1764.

WU Liang, HE Yi, MEI Xue, LIU Huan. Online behavior recognition using space-time interest points and probabilistic latent-dynamic conditional random field model[J]. Journal of Computer Applications, 2018, 38(6): 1760-1764.

参考文献

[1] 李瑞峰,王亮亮,王珂.人体动作行为识别研究综述[J].模式识别与人工智能,2014,27(1):35-48.(LI R F, WANG L L, WANG K. A survey of human body action recognition[J]. Pattern Recognition and Artificial Intelligence, 2014, 27(1):35-48.)
[2] 王亮.基于判别模式学习的人体行为识别方法研究[D].哈尔滨:哈尔滨工业大学,2011:56.(WANG L. Research on contrast pattern mining based human action recognition[D]. Harbin:Harbin Institute of Technology, 2011:56.)
[3] SUTTON C, MCCALLUM A. An introduction to conditional random fields[J]. Foundations and Trends in Machine Learning, 2011, 4(4):267-373.
[4] W LU, NG H T. Better punctuation prediction with dynamic conditional random fields[C]//Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing. Stroudsburg, PA:ACL, 2010:177-186.
[5] ZHANG J G, GONG S G. Action categorization with modified hidden conditional random field[J]. Pattern Recognition, 2010, 43(1):197-203.
[6] MAHMOUD E, AL-HAMADI A. LDCRFs-based hand gesture recognition[C]//Proceedings of 2012 IEEE International Conference on Systems, Man, and Cybernetics. Piscataway, NJ:IEEE, 2012:2670-2675.
[7] FUJⅡ Y, YAMAMOTO K, NAKAGAWA S. Hidden conditional neural fields for continuous phoneme speech recognition[J]. IEICE Transactions on Information and Systems, 2012, E95-D(8):2094-2104.
[8] van der MAATEN L, WELLING M, SAUL L K. Hidden unit conditional random fields[EB/OL].[2017-10-16]. https://staff.fnwi.uva.nl/m.welling/wp-content/uploads/papers/HiddenUnitCRF.pdf.
[9] 成立,梅雪,张玉燕,等.基于星形距离和LDCRF模型的在线行为识别方法[J]. 计算机工程与设计,2015,36(6):1626-1629.(CHENG L, MEI X, ZHANG Y Y, et al. Online behavior recognition based on star distance and LDCRF model[J]. Computer Engineering and Design, 2015, 36(6):1626-1629.)
[10] 张玉燕,梅雪,成立,等.基于隐动态条件神经域的在线行为识别方法[J].计算机工程与设计,2016,37(6):1632-1635.(ZHANG Y Y, MEI X, CHENG L, et al. Online behavior recognition based on latent-dynamic conditional neural fields[J]. Computer Engineering and Design, 2016, 37(6):1632-1635.)
[11] 马德智,李巴津,董志学.基于高斯混合模型的运动目标检测方法研究[J].电子测量技术,2013,36(10):47-50.(MA D Z, LI B J, DONG Z X. Moving object detection based on mixture Gaussian model[J]. Electronic Measurement Technology, 2013, 36(10):47-50.)
[12] LAPTEV I. On space-time interest points[J]. International Journal of Computer Vision, 2005, 64(2/3):107-123.
[13] PENG L, LU T W, MIN F. Multi-view indoor human behavior recognition based on 3D skeleton[C]//MIPPR 2015:Proceedings of the Ninth International Symposium on Multispectral Image Processing and Pattern Recognition, SPIE 9813. Bellingham:SPIE, 2015:98130Y.
[14] 祁家榕,张昌伟.行为识别技术的研究与发展[J].智能计算机与应用,2017,7(4):24-26.(QI J R, ZHANG C W. Research and development of behavior recognition technology[J]. Intelligent Computer and Applications, 2017, 7(4):24-26.)
[15] 赵雄伟.人体行为特征融合与行为识别的分析[J].无线互联科技,2017(12):104-105.(ZHAO X W. Analysis of human behavior characteristic fusion and behavior recognition[J]. Wireless Internet Technology, 2017(12):104-105.)

基于时空兴趣点和概率潜动态条件随机场模型的在线行为识别方法

Online behavior recognition using space-time interest points and probabilistic latent-dynamic conditional random field model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	李自强, 王正勇, 陈洪刚, 李林怡, 何小海. 基于外观和动作特征双预测模型的视频异常行为检测[J]. 计算机应用, 2021, 41(10): 2997-3003.
[2]	杨辉华, 张天宇, 李灵巧, 潘细朋. 基于MobileNet的移动端城管案件目标识别算法[J]. 计算机应用, 2019, 39(8): 2475-2479.
[3]	胡学敏, 易重辉, 陈钦, 陈茜, 陈龙. 基于运动显著图的人群异常行为检测[J]. 计算机应用, 2018, 38(4): 1164-1169.
[4]	杜启亮, 黎浩正, 田联房. 基于Adaboost和码本模型的手扶电梯出入口视频监控方法[J]. 计算机应用, 2017, 37(9): 2610-2616.
[5]	陈凯星, 刘赟, 王金海, 袁玉波. 基于遗传机制和高斯变差的自动前景提取方法[J]. 计算机应用, 2017, 37(11): 3231-3237.
[6]	姬丽娜, 陈庆奎, 陈圆金, 赵德玉, 方玉玲, 赵永涛. 基于GPU的视频流人群实时计数[J]. 计算机应用, 2017, 37(1): 145-152.
[7]	王智文, 蒋联源, 王宇航, 王日凤, 张灿龙, 黄镇谨, 王鹏涛. 基于尺度自适应局部时空特征的足球比赛视频中的多运动员行为表示[J]. 计算机应用, 2016, 36(8): 2134-2138.
[8]	潘磊, 周欢, 王明辉. 适用于密集人群的异常事件实时检测方法[J]. 计算机应用, 2016, 36(6): 1719-1723.
[9]	李慧慧, 秦品乐, 梁军. 基于HSI亮度分量和RGB空间的图像去雾算法[J]. 计算机应用, 2016, 36(5): 1378-1382.
[10]	王佩瑶, 曹江涛, 姬晓飞. 基于改进时空兴趣点特征的双人交互行为识别[J]. 计算机应用, 2016, 36(10): 2875-2879.
[11]	董小慧高戈陈亮韩镇江俊君. 数据驱动局部特征转换的噪声人脸幻构[J]. 计算机应用, 2014, 34(12): 3576-3579.
[12]	江爱文刘长红王明文. 基于前—后向光流点匹配运动熵的视频抖动检测算法[J]. 计算机应用, 2013, 33(10): 2918-2921.
[13]	谢璐金志刚王颖. 基于视频稳像和视角变换的公交客流计数方法[J]. 计算机应用, 2013, 33(10): 2926-2930.
[14]	杨泽平刘德强王茜向强铭. 基于地理信息系统技术的数据采集与监视控制系统设计[J]. 计算机应用, 2013, 33(02): 567-574.
[15]	周渝斌. 海量监控视频快速回放与检索技术[J]. 计算机应用, 2012, 32(11): 3185-3197.

基于时空兴趣点和概率潜动态条件随机场模型的 在线行为识别方法

Online behavior recognition using space-time interest points and probabilistic latent-dynamic conditional random field model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

基于时空兴趣点和概率潜动态条件随机场模型的在线行为识别方法