基于智能手机内置传感器的人体运动状态识别

doi:10.11959/j.issn.1000-436x.2019057

通信学报 ›› 2019, Vol. 40 ›› Issue (3): 157-169.doi: 10.11959/j.issn.1000-436x.2019057

基于智能手机内置传感器的人体运动状态识别

殷晓玲^1,²,陈晓江¹,夏启寿^1,²,何娟¹(),张鹏艳¹,陈峰¹

¹ 西北大学信息科学与技术学院，陕西西安 710127
² 池州学院数学与计算机学院，安徽池州 247000

修回日期:2018-12-11 出版日期:2019-03-01 发布日期:2019-04-04
作者简介:殷晓玲（1975- ），女，安徽枞阳人，池州学院副教授，主要研究方向为云计算、信息安全、机器学习。|陈晓江（1972- ），男，陕西西安人，博士，西北大学教授、博士生导师，主要研究方向为无线传感器网络定位、网络安全、软件体系结构。|夏启寿（1975- ），男，安徽庐江人，池州学院副教授，主要研究方向为云计算、信息安全、机器学习。|何娟（1994- ），女，江西萍乡人，西北大学硕士生，主要研究方向为无线传感器网络定位。|张鹏艳（1990- ）女，陕西西安人，西北大学硕士生，主要研究方向为无线传感器网络定位。|陈峰（1978- ），男，安徽天长人，博士，西北大学助理研究员，主要研究方向为无线网络。
基金资助:
国家自然科学基金资助项目(61170218);国家自然科学基金资助项目(61272461);国家自然科学基金资助项目(61373177)

Human motion state recognition based on smart phone built-in sensor

Xiaoling YIN^1,²,Xiaojiang CHEN¹,Qishou XIA^1,²,Juan HE¹(),Pengyan ZHANG¹,Feng CHEN¹

¹ School of Information Science and Technology,Northwest University,Xi’an 710127,China
² College of Mathematics and Computer Science,Chizhou University,Chizhou 247000,China

Revised:2018-12-11 Online:2019-03-01 Published:2019-04-04
Supported by:
The National Natural Science Foundation of China(61170218);The National Natural Science Foundation of China(61272461);The National Natural Science Foundation of China(61373177)

摘要/Abstract

摘要：

针对目前智能手机识别人体运动状态种类少、准确率低的问题，提出一种利用加速度传感器和重力传感器分层识别人体运动状态的方案。首先，利用加速度和重力加速度的关系计算出与手机方向无关的惯性坐标系下的线性加速度；其次，根据人体运动频率的变化范围和线性加速度矢量来确定脚步的波峰和波谷位置；最后，提取线性加速度在时域上的特征向量，使用层次支持向量机方法分层识别人体运动状态。实验结果表明，该方法能有效识别人体6种日常运动状态，准确率达到93.37%。

关键词: 运动状态识别, 层次支持向量机, 智能手机传感器, 时域特征

Abstract:

To solve problems of low accuracy and fewer types of human motion state recognized by current smart phones,a method to do hierarchical recognition by using acceleration sensors and gravity sensors was proposed.Firstly,linear acceleration in inertial coordinate system and independent of phone direction was calculated by using the relation between acceleration and gravity acceleration.Secondly,according to the span of human motion frequency and linear acceleration vector,positions of peak and trough of footsteps were determined.Finally,feature vector of linear acceleration in time domain was extracted and human motion states were recognized hierarchically by using hierarchical support vector machine (H-SVM).The experiment shows the method can recognize six usual human motion states,while accuracy rate up to 93.37%.

Key words: motion state recognition, hierarchical support vector machine, smart phone sensor,, time domain feature

中图分类号:

TP181

殷晓玲,陈晓江,夏启寿,何娟,张鹏艳,陈峰. 基于智能手机内置传感器的人体运动状态识别[J]. 通信学报, 2019, 40(3): 157-169.

Xiaoling YIN,Xiaojiang CHEN,Qishou XIA,Juan HE,Pengyan ZHANG,Feng CHEN. Human motion state recognition based on smart phone built-in sensor[J]. Journal on Communications, 2019, 40(3): 157-169.

图/表 31

图1

表1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

图12

图13

图14

图15

图16

图17

图18

图19

图20

图21

图22

表2

表3

图23

表4

表5

表6

表7

表8

参考文献 44

[1]	HENPRASERTTAE A , THIEMJARUS S , MARUKATAT S . Accurate activity recognition using a mobile phone regardless of device orientation and location[C]// International Conference on Body Sensor Networks. IEEE Computer Society, 2011: 41-46.
[2]	SHOAIB M , BOSCH S , INCEL O D ,et al. Fusion of smartphone motion sensors for physical activity recognition[J]. Sensors, 2014,14(6): 10146-10176.
[3]	GUO H , CHEN L , CHEN G ,et al. Smartphone‐based activity recog-nition independent of device orientation and placement[J]. Interna-tional Journal of Communication Systems, 2016,29(16): 2403-2415.
[4]	THIEMJARUS S , HENPRASERTTAE A , MARUKATAT S . A study on instance-based learning with reduced training prototypes for device-context-independent activity recognition on a mobile phone[C]// IEEE International Conference on Body Sensor Networks. IEEE, 2013: 1-6.
[5]	USTEV Y E , INCEL O D , ERSOY C . User,device and orientation independent human activity recognition on mobile phones:challenges and a proposal[C]// ACM Conference on Pervasive and Ubiquitous Computing Adjunct Publication. ACM, 2013: 1427-1436.
[6]	LU H , YANG J , LIU Z ,et al. The Jigsaw continuous sensing engine for mobile phone applications[C]// ACM Conference on Embedded Networked Sensor Systems. ACM, 2010: 71-84.
[7]	FIGO D , DINIZ P C , FERREIRA D R ,et al. Preprocessing techniques for context recognition from accelerometer data[J]. Personal ＆ Ubiquitous Computing, 2010,14(7): 645-662.
[8]	CAMHI S M , SISSON S B , JOHNSON W D ,et al. Accelerometer-determined moderate intensity lifestyle activity and cardiometabolic health[J]. Preventive Medicine, 2011,52(5): 358-360.
[9]	MIZELL D , . Using gravity to estimate accelerometer orientation[C]// IEEE International Symposium on Wearable Computers. IEEE, 2005: 252-253.
[10]	GUO H , CHEN L , CHEN G ,et al. Smartphone‐based activity recog-nition independent of device orientation and placement[J]. Interna-tional Journal of Communication Systems, 2016,29(16): 2403-2415.
[11]	李锋, 潘敬奎 . 基于三轴加速度传感器的人体运动识别[J]. 计算机研究与发展, 2016,53(3): 621-631.
	LI F , P J K . Human motion recognition based on triaxial accelerome-ter[J]. Journal of Computer Research and Development, 2016,53(3): 621-631.
[12]	CHEN Z , ZHU Q , CHAI S Y ,et al. Robust human activity recognition using smartphone sensors via CT-PCA and online SVM[J]. IEEE Transactions on Industrial Informatics, 2017,PP(99): 1-1.
[13]	李若涵, 张金艺, 徐德政 ,等. 运动分类步频调节的微机电惯性测量单元室内行人航迹推算[J]. 上海大学学报(自然科学版), 2014,20(5): 612-623.
	LI R H , ZHANG J Y , XU D Z ,et al. Micro-electro-mechanical sys-tem-inertial measurement unit indoor pedestrian dead reckoning based on motion classification and step frequency adjustment[J]. Journal of Shanghai University(Natural Science), 2014,20(5): 612-623.
[14]	RAVI N , DANDEKAR N , MYSORE P ,et al. Activity recognition from accelerometer data[C]// The Twentieth National Conference on Artificial Intelligence and the Seventeenth Innovative Applications of Artificial Intelligence Conference. DBLP, 2005: 1541-1546.
[15]	KASTEREN T V , NOULAS A , ENGLEBIENNE G . Accurate activityrecognition in a home setting[C]// International Conference on Ubiquitous Computing. ACM, 2008: 1-9.
[16]	CASALE P , PUJOL O , RADEVA P . Human activity recognition from accelerometer data using a wearable device[M]// Pattern Recognition and Image Analysis. Springer Berlin Heidelberg, 2011: 289-296.
[17]	EMMANUEL M T . Using machine learning for real-time activity recognition and estimation of energy expenditure[D]. Bosto :Massachusetts Institute of Technology, 2008: 50-57.
[18]	KHAN A M , LEE Y K , LEE S Y ,et al. A triaxial accelerometer-based physical-activity recognition via augmented-signal features and a hierarchical recognizer[J]. IEEE Transactions on Information Technology in Biomedicine A Publication of the IEEE Engineering in Medicine ＆Biology Society, 2010,14(5): 1166-1172.
[19]	KWAPISZ J R , WEISS G M , MOORE S A . Activity recognition using cell phone accelerometers[J]. ACM SIGKDD Explorations Newsletter, 2011,12(2): 74-82.
[20]	SIIRTOLA P , JUHA R?NING . Recognizing human activities user-independently on smartphones based on accelerometer data[J]. International Journal of Interactive Multimedia ＆ Artificial Intelligence, 2012,1(5): 38-45.
[21]	HE Y , LI Y , BAO S D . Fall detection by built-in tri-accelerometer of smartphone[C]// IEEE-EMBS International Conference on Biomedical and Health Informatics. IEEE, 2012: 184-187.
[22]	USTEV Y E , DURMAZ INCEL O , ERSOY C . User,device and orientation independent human activity recognition on mobile phones:challenges and a proposal[C]. The 2013 ACM Conference On Pervasive And Ubiquitous Computing Adjunct Publication. ACM, 2013: 1427-1436.
[23]	MITCHELL E , MONAGHAN D , O'CONNOR N E . Classification of sporting activities using smartphone accelerometers[J]. Sensors, 2013,13(4): 5317-5337.
[24]	KHAN A M , SIDDIQI M H , LEE S W . Exploratory data analysis of acceleration signals to select light-weight and accurate features for real-time activity recognition on smartphones[J]. Sensors, 2013,13(10): 99-122.
[25]	ZHANG L , WU X , LUO D . Improving activity recognition with context information[C]// IEEE International Conference on Mechatronics and Automation. IEEE, 2015: 1241-1246.
[26]	WANG A , CHEN G , YANG J ,et al. A comparative study on human activity recognition using inertial sensors in a smartphone[J]. IEEE Sensors Journal, 2016,16(11): 4566-4578.
[27]	SHOAIB M , BOSCH S , INCEL O D ,et al. Complex activity recognition using smartphone and wrist-worn motion sensors[J]. Sensors, 2016,16(4):426.
[28]	JONGPRASITHPORN M , YODPIJIT N , SRIVILAI R ,et al. A smartphone-based real-time simple activity recognition[C]// International Conference on Control,Automation and Robotics. IEEE, 2017: 539-542.
[29]	CHEN Y , SHEN C . Performance Analysis of Smartphone-Sensor Behavior for Human Activity Recognition[J]. IEEE Access, 2017,5(99): 3095-3110.
[30]	彭际群 . 基于加速度传感器的人体运动状态识别研究[D]. 哈尔滨:哈尔滨工业大学, 2014.
	PENG J Q . The research of human activity state recognition base on accelerometers[D]. Harbin :Harbin Institute of Technology, 2014
[31]	GLUECK M , BUHMANN A , MANOLI Y . Autocalibration of MEMSaccelerometers[C]// Instrumentation and Measurement Technology Conference. IEEE, 2012: 1788-1793.
[32]	陆欣, 刘忠, 张宏欣 ,等. 三轴MEMS加速度计的最大似然校正算法[J]. 国防科技大学学报, 2017,39(5): 185-191.
	LU X , LIU ZH , ZHANG H X ,et al. Maximum likelihood calibration for MEMS triaxial accelerometer[J]. Journal of National University of Defense Technology, 39(5): 185-191.
[33]	杨管金子, 李建辰, 黄海 ,等. 基于遗传算法的加速度计免转台标定方法[J]. 中国惯性技术学报, 2017,25(1): 119-123.
	YANG G J Z , LI J C , HUANG H ,et al. Non-turntable calibration method for three-accelerometer based on genetic algorithm[J]. Journal of Chinese Inertial Technology, 2017,25(1): 119-123.
[34]	毛友泽, 张海 . 无依托状态加速度计的新型标定方法[J]. 北京航空航天大学学报, 2011,37(1): 123-126.
	MAO Y Z , ZHANG H . Novel Calibration method for accelerometers of in-field use[J]. Journal of Beijing University of Aeronautics and As-tronautics, 2011,37(1): 123-126.
[35]	桂毅, . 基于高斯牛顿迭代法的 MEMS 加速度计标定方法[C]// 2016 IEEE Chinese Guidance,Navigation and Control Conference. 2016: 3032-3036.
	GUI Y , . Calibration of MEMS accelerometer based on gauss-newton iteration algorithm[C]// 2016 IEEE Chinese Guidance,Navigation and Control Conference. 2016: 3032-3036.
[36]	戴邵武, 王克红, 庆曾宏 . 基于混合 PSO 算法的加速度计快速标定方法[J]. 压电与声光, 2015,37(2): 232-236.
	DAI S W , WANG K H , Qing Z H . Rapid calibration of accelerometer based on hybrid particle swarm optimization algorithm[J]. Piezoelectrics ＆ Acoustooptics, 2015,37(2): 232-236.
[37]	KRISHNAN N C , COOK D J . Activity recognition on streaming sensor data[J]. Pervasive ＆ Mobile Computing, 2014,10(Pt B): 138-154.
[38]	BANOS O , GALVEZ J M , DAMAS M ,et al. Window size impact in human activity recognition[J]. Sensors, 2014,14(4): 6474-6499.
[39]	ROBNIK , IKONJA M , KONONENKO I . Theoretical and empirical analysis of ReliefF and RReliefF[J]. Machine Learning, 2003,53(1-2): 23-69.
[40]	LIU J , CHEN R , LING P ,et al. A hybrid smartphone indoor positioning solution for mobile LBS[J]. Sensors, 2012,12(12): 17208-17233.
[41]	梁久祯, 朱向军, 陈璟 . 基于手机加速度传感器的高精低采样计步算法设计[J]. 西北大学学报(自然科学版), 2015,45(5): 738-744.
	LIANG J ZH , ZHU X J , CHEN J . A high-accuracy and low-sampling-rate step counting algorithm design based on smartphone’s accelerometer[J]. Journal of Northwest Universi-ty(Natural Science Edition), 2015,45(5): 738-744.
[42]	凌海波, 杨静, 周先存 . 基于手机加速度传感器的波峰波谷计步算法研究[J]. 四川理工学院学报(自然科学版), 2017,30(3): 21-25.
	LING H B , YANG J , ZHOU X C . Research on peak and trough step counting algorithm based on mobile accelerometer[J]. Journal of Si-chuan University of Science ＆ Engineering(Natural Science Edi-tion), 2017,30(3): 21-25.
[43]	FIGO D , DINIZ P C , FERREIRA D R ,et al. Preprocessing techniques for context recognition from accelerometer data[J]. Personal ＆ Ubiquitous Computing, 2010,14(7): 645-662.
[44]	ANGUITA D , GHIO A , ONETO L ,et al. Human activity recognition on smartphones using a multiclass hardware-friendly support vector machine[C]// International Conference on Ambient Assisted Living and Home Care. Springer-Verlag, 2012: 216-223.

型号	CPU	内存/GB
华为mate 9	Kirin 960 8核2.4 GHz	6
小米 max2	MSM8953 8核2.0 GHz	4
OPPO A73	MT6763T 8核2.5 GHz	4
vivo X9s	MSM8976 8核1.8 GHz	4
三星C5	MSM8953pro 8核2.2 GHz	4

状态	胸口	上臂	裤前袋	裤后袋	腰部	综合
静止	97.98%	90.27%	96.59%	96.00%	98.92%	98.52%
骑行	96.39%	96.63%	92.69%	95.43%	98.62%	97.52%
跑步	96.66%	89.60%	93.51%	96.64%	97.58%	96.71%
行走	86.38%	83.29%	86.05%	86.38%	90.91%	90.03%
上楼	86.55%	82.76%	85.56%	84.59%	89.83%	88.95%
下楼	86.03%	83.22%	84.55%	84.03%	89.28%	88.51%
均值	91.66%	87.63%	89.82%	90.51%	94.19%	93.37%

状态	静止	骑行	跑步	行走	上楼	下楼
静止	7 403	92	0	12	4	3
骑行	120	8 141	10	70	3	4
跑步	0	15	11 848	218	89	81
行走	5	8	165	9 482	448	424
上楼	2	5	35	578	8 755	468
下楼	2	7	28	595	495	8 682

手机放置方式			DE-PSO算法			高斯-牛顿迭代法
X轴	Y轴	Z轴	X轴	Y轴	Z轴	X轴	Y轴	Z轴
9.810	0.000	0.000	9.794	0.015	0.015	9.733	0.040	0.049
-9.810	0.000	0.000	-9.792	0.016	0.015	-9.844	0.049	0.042
0.000	9.810	0.000	0.017	9.796	0.013	0.030	9.846	0.044
0.000	-9.810	0.000	0.016	-9.788	0.011	0.048	-9.738	0.032
0.000	0.000	9.810	0.016	0.016	9.813	0.035	0.034	9.745
0.000	0.000	-9.810	0.015	0.043	-9.815	0.049	0.049	-9.739

方法	校准前	校准后
文献[9]	90.12%	91.83%
文献[10]	91.81%	92.41%
文献[11]	90.13%	91.87%
文献[12]	91.72%	92.33%
本文	92.21%	93.37%

基于智能手机内置传感器的人体运动状态识别

Human motion state recognition based on smart phone built-in sensor

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 31

参考文献 44

相关文章 15

Metrics

推荐阅读 0

位置	文献[41]	文献[42]	本文
手持(上臂)	97.00%	93.00%	96.75%
裤兜(前后)	92.00%	89.00%	97.58%
背包(腰部)	93.00%	89.50%	98.00%
上衣(胸口)	93.00%	—	97.25%
均值	93.75%	90.50%	97.39%

方法	位置数	运动状态数	特征向量	计算复杂度	准确率
文献[11]	1	6	时域、频域	O(nlbn)	96.10%
文献[12]	3	5	时域、频域	O(nlbn)	90.44%
文献[19]	1	6	时域	O(n)	91.70%
文献[44]	1	6	时域、频域	O(nlbn)	89.00%
本文	5	6	时域	O(n)	93.49%

方法	识别率
文献[11]	90.35%
文献[12]	91.28%
文献[19]	85.76%
文献[44]	84.74%
本文	93.49%

[1]	钱榕, 许建婷, 张克君, 董宏宇, 邢方远. 隐马尔可夫模型的异质网络链接预测方法研究[J]. 通信学报, 2022, 43(5): 214-225.
[2]	顾亦然, 姚朱鹏, 杨海根. 融合注意力胶囊的深度因子分解机模型[J]. 通信学报, 2021, 42(10): 130-139.
[3]	顾秋阳, 吴宝, 孙兆洋, 池仁勇. 基于改进灰狼优化的复杂网络重要节点识别算法[J]. 通信学报, 2021, 42(6): 72-83.
[4]	杨晓晖,刘晓明. 基于双向邻居修正的局部异常因子算法[J]. 通信学报, 2020, 41(8): 130-140.
[5]	徐丰力,李勇. 城市环境下的用户移动行为建模概述[J]. 通信学报, 2020, 41(7): 18-28.
[6]	顾纯祥,吴伟森,石雅男,李光松. 基于自编码器的未知协议分类方法[J]. 通信学报, 2020, 41(6): 88-97.
[7]	邱飞岳,陈博文,陈铁明,章国道. 稀疏诱导流形正则化凸非负矩阵分解算法[J]. 通信学报, 2020, 41(5): 84-95.
[8]	闫光辉, 张萌, 罗浩, 李世魁, 刘婷. 融合高阶信息的社交网络重要节点识别算法[J]. 通信学报, 2019, 40(10): 109-118.
[9]	吴宾,陈允,孙中川,叶阳东. 联合成对排序的物品推荐模型[J]. 通信学报, 2019, 40(9): 193-206.
[10]	杨晓晖,张圣昌. 基于多粒度级联孤立森林算法的异常检测模型[J]. 通信学报, 2019, 40(8): 133-142.
[11]	刘浩然,丁攀,郭长江,常金凤,崔静闯. 基于贝叶斯算法的中文垃圾邮件过滤系统研究[J]. 通信学报, 2018, 39(12): 151-159.
[12]	张俐,王枞. 基于最大相关最小冗余联合互信息的多标签特征选择算法[J]. 通信学报, 2018, 39(5): 111-122.
[13]	李幼军,黄佳进,王海渊,钟宁. 基于SAE和LSTM RNN的多模态生理信号融合和情感识别研究[J]. 通信学报, 2017, 38(12): 109-120.
[14]	康岚兰,董文永,宋婉娟,李康顺. 无惯性自适应精英变异反向粒子群忧化算法[J]. 通信学报, 2017, 38(8): 66-78.
[15]	章鹏,刘全,钟珊,翟建伟,钱炜晟. 增量式双自然策略梯度的行动者评论家算法[J]. 通信学报, 2017, 38(4): 166-177.