首页 > 论文 > 光学学报 > 34卷 > 10期(pp:1015006--1)

基于局部轮廓和随机森林的人体行为识别

Human Action Recognition Based on Local Image Contour and Random Forest

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

基于视频信息的人体行为识别得到了越来越多的关注。针对人体行为的局部表达,提出了一种新的局部轮廓特征来描述人体的外观姿势,可以同时利用水平和竖直方向上的轮廓变化信息。该特征能有效区分不同动作,与轮廓起始点无关,具有平移、尺度和旋转不变性。针对该特征,提出了一种基于随机森林的两阶段分类方法,使用随机森林分类器对行为视频的局部轮廓进行初分类,并根据每个局部轮廓对应决策类的分类树数目占总分类树数目的比例,提出了一种基于袋外(OOB)数据误差加权投票准则的行为视频分类算法。在测试数据集上的实验结果证实了该方法的有效性。

Abstract

Human action recognition in videos has attracted more and more attentions. In view of the local expression of human behavior, a novel local contour feature representing body posture is proposed, which can make full use of information of the contour variation along both horizontal and vertical direction. The proposed local feature can distinguish different actions and is invariant to translation, scaling, rotation and change of start point of human contour. A two stage classifying framework based on random forest is also proposed by using this novel local body contour feature. Random forest is employed to classify each frame of the test video. After that, a video classification method based on out of bag(OOB) error weighted voting strategy to recognize action video according to the ratio of decision trees belonging to each local contour to total decision trees is proposed. Experimental results on test data set prove the effectiveness of proposed method.

Newport宣传-MKS新实验室计划
补充资料

中图分类号:TP391

DOI:10.3788/aos201434.1015006

所属栏目:机器视觉

基金项目:国家自然科学基金(61272338)

收稿日期:2014-03-26

修改稿日期:2014-06-23

网络出版日期:--

作者单位    点击查看

蔡加欣:中山大学数学与计算科学学院, 广东 广州 510275广东省计算科学重点实验室, 广东 广州 510275
冯国灿:中山大学数学与计算科学学院, 广东 广州 510275广东省计算科学重点实验室, 广东 广州 510275
汤鑫:中山大学数学与计算科学学院, 广东 广州 510275广东省计算科学重点实验室, 广东 广州 510275
罗志宏:中山大学信息科学与技术科学学院, 广东 广州 510275

联系人作者:蔡加欣(caijxin@mail2.sysu.edu.cn)

备注:蔡加欣(1988—),男,博士研究生,主要从事机器视觉和医学图像处理方面的研究。

【1】Li Hongsong, Li Da. Some advances in human motion analysis [J]. PR&AI, 2009, 22(1): 70-78.
黎洪松, 李达. 人体运动分析研究的若干新进展[J]. 模式识别与人工智能, 2009, 22(1): 70-78.

【2】Xu Guangyou, Cao Yuanyuan. Action recognition and activity understanding: a review [J]. Journal of Image and Graphics, 2009, 14(2): 189-195.
徐光祐, 曹媛媛. 动作识别与行为理解综述[J]. 中国图象图形学报, 2009, 14(2): 189-195.

【3】A F Bobick, J W Davis. The recognition of human movement using temporal templates [J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2001, 23(3): 257-267.

【4】D Weinland, R Ronfard, E Boyer. Free viewpoint action recognition using motion history volumes [J]. Computer Vision and Image Understanding, 2006, 104(2-3): 249-257.

【5】J X Cai, G C Feng, X Tang. Human action recognition using oriented holistic feature [C]. 20th IEEE International Conference on Image Processing, 2013. 2420-2424.

【6】Guan Zhiqiang, Chen Qian, Gu Guohua, et al.. Dim target detection based on optical flow histgram in low frame frequence in clouds background [J]. Acta Optica Sinica, 2008, 28(8): 1496-1501.
管志强, 陈钱, 顾国华, 等. 基于光流直方图的云背景下低帧频小目标探测方法[J]. 光学学报, 2008, 28(8): 1496-1501.

【7】Wang Xiangjun, Wang Yan, Li Zhi. Fast target recognition and tracking method based on characteristic corner [J]. Acta Optica Sinica, 2007, 27(2): 360-364.
王向军, 王研, 李智. 基于特征角点的目标跟踪和快速识别算法研究[J]. 光学学报, 2007, 27(2): 360-364.

【8】Gao Lin, Tang Peng, Sheng Peng, et al.. Visual object tracking based on conditional random field under complex scene [J]. Acta Optica Sinica, 2010, 30(6): 1721-1728.
高琳, 唐鹏, 盛鹏, 等. 复杂场景下基于条件随机场的视觉目标跟踪[J]. 光学学报, 2010, 30(6): 1721-1728。

【9】Jin Biao, Hu Wenlong, Wang Hongqi. Moving-objects interaction recognition based on the spatial-temporal semantic information [J]. Acta Optica Sinica, 2012, 32(5): 0515002.
金标, 胡文龙, 王宏琦. 基于时空语义信息的视频运动目标交互行为识别方法[J]. 光学学报, 2012, 32(5): 0515002.

【10】J C Niebles, H C Wang, F F Li. Unsupervised learning of human action categories using spatial-temporal words [J]. Int J Comput Vis, 2008, (79): 299-318.

【11】K Alexander, M Marcin, S Cordelia. A spatio-temporal descriptor based on 3d-gradients [C]. British Machine Vision Conference, IEEE Computer Society, 2008. 995-1004.

【12】S Poul, A Saad, S Mubarak. A 3-dimensional sift descriptor and its application to action recognition [C]. Proceedings of the 15th international conference on Multimedia, IEEE Computer Society, 2007. 357-360.

【13】Haihua Cui, Wenhe Liao, Ning Dai, et al.. Registration and integration algorithm in structured light three-dimensional scanning based on scale-invariant feature matching of multi-source images [J]. Chin Opt Lett, 2012, 10(9): 091001.

【14】Jianfang Dou, Jianxun Li. Robust image matching based on SIFT and delaunay triangulation [J]. Chin Opt Lett, 2012, 10(s1): s11001.

【15】Mingliang Gao, Xiaomin Yang, Yanmei Yu, et al.. Photometric invariant feature descriptor based on SIFT [J]. Chin Opt Lett, 2012, 10(s1): s11003.

【16】J G Liu, S Ali, M Shah. Recognizing human actions using multiple features [C]. Computer Vision and Pattern Recognition, IEEE Computer Society, 2008. 1-8.

【17】J Yamato, J Ohya, K Ishii. Recognizing human action in time sequential images using hidden Markov model[C]. IEEE Conference on Computer Vis Pattern Recognit, 1992. 379-385.

【18】M Brand, N Oliver, A Pentland. Coupled hidden Markov models for complex action recognition [C]. IEEE Conference on Computer Vis Pattern Recognit, 1997: 994-999.

【19】Q F Shi, L Wang, L Cheng, et al.. Discriminative human action segmentation and recognition using semi-markov model [C]. IEEE Comference on Computer Vision and Pattern Recognition, 2008: 1-5.

【20】C Sminchisescu, A Kanaujia, D Metaxax. Conditional models for contextual human motion recognition [J]. Computer Vision and Jmago Understanding, 2006, 104(2-3): 210-220.

【21】L Wang, D Suter. Visual learning and recognition of sequential data manifolds with applications to human movement analysis [J]. Computer Vision and Image Understanding, 2008, 110(2): 153-172.

【22】S Cheema, A Eweiwi, C Thurau, et al.. Action recognition by learning discriminative key poses [C]. IEEE International Conference on Computer Vision Workshops, 2011. 1302-1309.

【23】C A Andre, C P Pau, Flórez-Revuelta Francisco. Silhouette-based human action recognition using sequences of key poses [J]. Pattern Recognition Letters, 2013, 34(15): 1799-1807.

【24】Ming Ying, Jiang Jingjue. Background modeling and moving-objects detection based on cauchy distribution for video sequence [J]. Acta Optica Sinica, 2008, 28(3): 587-592.
明英, 蒋晶珏. 基于柯西分布的视频图像序列背景建模和运动目标检测[J]. 光学学报, 2008, 28(3): 587-592.

【25】S Suzuki, K Be. Topological structural analysis of digitized binary images by border following[J]. Comput Vision Graphics Image Process, 1985, 30(1): 32-46.

【26】H T Kam. The random subspace method for constructing decision forests [J]. Pattern Analysis and Machine, 1998, 20(8): 832-844.

【27】B Leo. Random Forests [J]. Machine Learning, 2001, 45(1): 5-32.

【28】Zhang Hongqiang, Liu Guangyuan, Lai Xiangwei. Application of random forest algorithm in important feature selection from EMG signal [J]. Computer Science, 2013, 40(1): 200-202.
张洪强, 刘光远, 赖祥伟. 随机森林算法在肌电的重要特征选择中的应用[J]. 计算机科学, 2013, 40(1): 200-202.

【29】Guo Jinxin, Chen Wei. Face recognition based on HOG multi-feature fusion and random forest [J]. Computer Science, 2013, 40(10): 279-282.
郭金鑫, 陈玮. 基于HOG多特征融合与随机森林的人脸识别[J]. 计算机科学, 2013, 40(10): 279-282.

【30】Haixia Xu, Xianbin Wen, Yongliao Zou, et al.. Performance evaluation and segmentation for synthetic aperture radar image using mixture multiscale autoregressive model and bootstrap technique [J]. Chin Opt Lett, 2012, 10(s1): s11005.

【31】B Mosh, G Lena, S Eli, et al.. Actions as space-time Shapes [C]. IEEE International Conference on Computer Vision, 2005, 2: 1395-1402.

【32】Q Zhao, H S Horace. Unsupervised approximate-semantic vocabulary learning for human action and video classification [J]. Pattern Recognition Letters, 2013, 34(15): 1870-1878.

【33】S Singh, S Velatin, H Ragheb. Muhavi: a multicamera human action video dataset for the evalution of action recognition methods [C]. IEEE International Conference on Boston: Adavanced Video and Signal Based Surveillance(AVSS), 2010. 48-55.

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF