基于隐马尔科夫模型的人体动作压缩红外分类

关秋菊; 罗晓牧; 郭雪梅; 王国利

doi:10.16383/j.aas.2017.c160130

基于隐马尔科夫模型的人体动作压缩红外分类

doi: 10.16383/j.aas.2017.c160130

1.
仲恺农业工程学院机电工程学院广州 510225
2.
广州中医药大学医学信息工程学院广州 510006
3.
中山大学数据科学与计算机学院广州 510006

基金项目:

广东省教育厅青年创新人才项目 2015KQNCX068

国家自然科学基金 61375080

国家自然科学基金 61601523

国家自然科学基金 61301294

广东省自然科学基金 2015A030311049

广东省自然科学基金 2016A030310238

详细信息

作者简介:
关秋菊仲恺农业工程学院机电工程学院讲师.主要研究方向为基于周边智能环境的人体运动行为识别.E-mail:qiuju95@126.com

罗晓牧广州中医药大学医学信息工程学院讲师.2012年获得中山大学通信与信息系统专业博士学位.主要研究方向为无线传感器网络, 人体动作识别和机器学习.E-mail:woodwood2000@163.com

郭雪梅中山大学数据科学与计算机学院副教授.主要研究方向为嵌入式系统与移动计算, 信息获取与信息处理.E-mail:guoxuem@mail.sysu.edu.cn

计量
- 文章访问数: 1964
- HTML全文浏览量: 287
- PDF下载量: 914
- 被引次数: 0
出版历程
- 收稿日期: 2016-02-03
- 录用日期: 2016-09-30
- 刊出日期: 2017-03-20

Compressive Infrared Classification of Human Motion Using HMM

1.
College of Mechanical and Electrical Engineering, Zhongkai University of Agriculture Engineering, Guangzhou 510225
2.
School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou 510006
3.
School of Data and Computer Science, Sun Yat-Sen University, Guangzhou 510006

Funds:

Young Creative Talents Project of Guangdong Province Education Department 2015KQNCX068

National Natural Science Foundation of China 61375080

National Natural Science Foundation of China 61601523

National Natural Science Foundation of China 61301294

Natural Science Foundation of Guangdong Province 2015A030311049

Natural Science Foundation of Guangdong Province 2016A030310238

More Information

Author Bio:
Lecturer at the College of Mechanical and Electrical Engineering, Zhongkai University of Agriculture and Engineering. Her main research interest is physical activity recognition based on ambient intelligence environment

Lecturer at the School of Medical Information Engineering, Guangzhou University of Chinese Medicine. He received his Ph. D. degree from Sun Yat-Sen University in 2012. His research interest covers wireless sensor networks, human activity recognition and machine learning

Associate professor at the School of Data and Computer Science, Sun Yat-Sen University. Her research interest covers embedded system, mobile computing, information acquisition and processing

Corresponding author: WANG Guo-LiProfessor at the School of Data and Computer Science, Sun Yat-Sen University. His research interest covers ambient intelligence and ambient assisted human-robot collaboration. Corresponding author of this paper

摘要

摘要: 人体动作产生的辐射能量变化（Infrared radiation changes，IRC）信号是动作识别的重要线索，本文提出了一种基于隐马尔科夫模型的人体动作压缩红外分类新方法.针对人体动作的自遮挡问题，建立基于正交视角的压缩红外测量系统，获取人体动作在主投影面和辅助投影面的IRC压缩信号；然后，采用隐马尔科夫模型（Hidden Markov model，HMM）双层特征建模算法进行压缩域动作分类.实验结果表明双层特征建模的平均正确分类率高于主层特征建模，平均正确分类率可达95.71%.该方法为环境辅助生活系统提供了人体动作识别的新途径.
- 环境辅助生活 /
- 隐马尔科夫模型 /
- 压缩感知 /
- 热释电红外传感器 /
- 动作分类
Abstract: Infrared radiation changes (IRC) induced by human motion can provide important clue for motion classification. This paper presents a hidden Markov model (HMM)-based compressive infrared classification method to recognize human motions. In order to solve the problem of self-occlusion, an orthogonal-view based compressive infrared sensing system is implemented for projecting the IRC to two orthogonal planes in the infrared radiation field. Then, a double-layer feature model using HMM classifier is trained to carry out motion recognition with the compressive measurements. Experimental results show that the mean correct classification rate with double-layer feature is 95.71%, which is better than that with main-layer feature. This method provides a new approach to classification of human motions for ambient assisted system.
- Ambient assisted living (AAL) /
- hidden Markov model (HMM) /
- compressive sensing /
- pyroelectric infrared sensors /
- motion classification
注释:

1) 本文责任编委王启宁

HTML全文

图 1 红外动作识别实验场景

Fig. 1 The experiment scenario for infrared motion classification

下载: 全尺寸图片幻灯片

图 2 人体动作压缩红外测量信号 (“m3挥双臂"动作)

Fig. 2 The PIR sensors output recorded for the motion of "waving two hands" (m3)

下载: 全尺寸图片幻灯片

图 3 GMHMM建模中隐状态数目、高斯模型数目对平均对数似然值的影响 (动作m1)

Fig. 3 The influences of the HMM parameters on the average log likelihood (motion m1)

下载: 全尺寸图片幻灯片

图 4 动作m9的一个测试样本在主层GMHMM下的对数似然概率

Fig. 4 Log likelihoods of a test sequence of motion m9 using main-layer GMHMM

下载: 全尺寸图片幻灯片

图 5 动作m9的一个测试样本在双层GMHMM下的对数似然概率

Fig. 5 Log likelihoods of a test sequence of motion m9 using double-layer GMHMM

下载: 全尺寸图片幻灯片

图 6 测量维数对正确分类率的影响

Fig. 6 The influence of the measurement dimension on the CCR

下载: 全尺寸图片幻灯片

图 7 测量维数对分类时间的影响

Fig. 7 The influence of the measurement dimension on the running time

下载: 全尺寸图片幻灯片

图 8 测量视角对正确识别率的影响

Fig. 8 The influences of the sensing view on the CCR

下载: 全尺寸图片幻灯片

表 1 GMHMMs参数配置

Table 1 The specification of GMHMMs

特征层模型	m1		m2		m3		m4		m5		m6		m7		m8		m9		m10
	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$	$M_G$	$H$
主层GMHMM	2	5	2	7	2	9	4	3	2	7	2	6	2	5	2	3	2	5	2	6
双层GMHMM	2	9	2	4	2	5	3	3	2	8	2	5	2	4	2	4	2	3	2	5

下载: 导出CSV

表 2 测量视角对CCR的影响

Table 2 Sensing view vs. CCR

测量视角	正视	顶视	正交视角
CCR (%)	94.81	87.23	95.71

下载: 导出CSV

参考文献(23)

[1]	Wichert R, Eberhardt B. Ambient assisted living. Advanced Technologies & Societal Change. Berlin, Germany: Springer-Verlag Berlin Heidelberg, 2012. 1145-1148
[2]	Chen L M, Hoey J, Nugent C D, Cook D J, Yu Z W. Sensor-based activity recognition. IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), 2012, 42(6): 790-808 doi: 10.1109/TSMCC.2012.2198883
[3]	Aggarwal J K, Ryoo M S. Human activity analysis: a review. ACM Computing Surveys, 2011, 43(3): 16 http://www.bibsonomy.org/bibtex/248f6b014794b15388f00a8bad140cb56/flint63
[4]	Hu W M, Tan T N, Wang L, Maybank S. A survey on visual surveillance of object motion and behaviors. IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), 2004, 34(3): 334-352 doi: 10.1109/TSMCC.2004.829274
[5]	Moeslund T B, Hilton A, Krüger V. A survey of advances in vision-based human motion capture and analysis. Computer Vision and Image Understanding, 2006, 104(2-3): 90-126 doi: 10.1016/j.cviu.2006.08.002
[6]	Turaga P, Chellappa R, Subrahmanian V S, Udrea O. Machine recognition of human activities: a survey. IEEE Transactions on Circuits and Systems for Video Technology, 2008, 18(11): 1473-1488 doi: 10.1109/TCSVT.2008.2005594
[7]	Johansson G. Visual perception of biological motion and a model for its analysis. Perception & Psychophysics, 1973, 14(2): 201-211 http://www.worldcat.org/title/visual-perception-of-biological-motion-and-a-model-for-its-analysis/oclc/470021412
[8]	Shi G Y, Chan C S, Li W J, Leung K S, Zou Y X, Jin Y F. Mobile human airbag system for fall protection using MEMS sensors and embedded SVM classifier. IEEE Sensors Journal, 2009, 9(5): 495-503 doi: 10.1109/JSEN.2008.2012212
[9]	Yang A Y, Jafari R, Sastry S S, Bajcsy R. Distributed recognition of human actions using wearable motion sensor networks. Journal of Ambient Intelligence and Smart Environments, 2009, 1(2): 103-115
[10]	Burchett J, Shankar M, Hamza A B, Guenther B D, Pitsianis N, Brady D J. Lightweight biometric detection system for human classification using pyroelectric infrared detectors. Applied Optics, 2006, 45(13): 3031-3037 doi: 10.1364/AO.45.003031
[11]	Candés E J, Wakin M B. Wakin. An introduction to compressive sampling. IEEE Signal Processing Magazine, 2008, 25(2): 21-30 doi: 10.1109/MSP.2007.914731
[12]	Duarte D F, Eldar Y C. Structured compressed sensing: from theory to applications. IEEE Transactions on Signal Processing, 2011, 59(9): 4053-4085 doi: 10.1109/TSP.2011.2161982
[13]	Brady D J, Pitsianis N P, Sun X B. Reference structure tomography. Journal of the Optical Society of America A, 2004, 21(7): 1140-1147 doi: 10.1364/JOSAA.21.001140
[14]	Peng M, Xiao Y. A survey of reference structure for sensor systems. IEEE Communications Surveys & Tutorials, 2012, 14(3): 897-910 https://www.researchgate.net/publication/224256688_A_Survey_of_Reference_Structure_for_Sensor_Systems
[15]	Wimalajeewa T, Chen H, Varshney P K. Performance limits of compressive sensing-based signal classification. IEEE Transactions on Signal Processing, 2012, 60(6): 2758-2770 doi: 10.1109/TSP.2012.2189859
[16]	Davenport M A, Boufounos P T, Wakin M B, Baraniuk R G. Signal processing with compressive measurements. IEEE Journal of Selected Topics in Signal Processing, 2010, 4(2): 445-460 doi: 10.1109/JSTSP.2009.2039178
[17]	Luo X M, Liu T, Liu J, Guo X M, Wang G L. Design and implementation of a distributed fall detection system based on wireless sensor networks. EURASIP Journal on Wireless Communications and Networking, 2012, 2012(1): 118 doi: 10.1186/1687-1499-2012-118
[18]	Sun Q Q, Hu F, Hao Q. Mobile target scenario recognition via low-cost pyroelectric sensing system: Toward a context-enhanced accurate identification. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2014, 44(3): 375-384 doi: 10.1109/TSMC.2013.2263130
[19]	Guan Q J, Li C Y, Guo X M, Wang G L. Compressive classification of human motion using pyroelectric infrared sensors. Pattern Recognition Letters, 2014, 49: 231-237 doi: 10.1016/j.patrec.2014.07.018
[20]	Kay S M. Fundamentals of Statistical Signal Processing: Detection Theory. Englewood Cliffs, NJ: Prentice-Hall, 1998.
[21]	Gales M J F. Maximum likelihood linear transformations for HMM-based speech recognition. Computer Speech & Language, 1998, 12(2): 75-98 http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.21.444
[22]	Babu R V, Anantharaman B, Ramakrishnan K R, Srinivasan S H. Compressed domain action classification using HMM. Pattern Recognition Letters, 2002, 23(10): 1203-1213 doi: 10.1016/S0167-8655(02)00067-3
[23]	Li H, Greenspan M. Model-based segmentation and recognition of dynamic gestures in continuous video streams. Pattern Recognition, 2011, 44(8): 1614-1628 doi: 10.1016/j.patcog.2010.12.014