A Detection Method for the Fall Behavior of Elders Based on Hidden Markov Model
Cao Huiqiang1, Lin Chungchih2, Wu Shuicai1*
1(College of Life Science and Bio-engineering, Beijing University of Technology, Beijing 100124,China) 2(Department of Computer Science and Information Engineering, Chang Gung University,Taoyuan 33302, Taiwan, China)
Abstract：As the social aging process quickened, the demand to care the elderly's health and safety is increasing. The fall in the elderly population is a very common phenomenon; it has been a major health risk that diminishes the quality of life among the elderly people. In this paper, we proposed a new method using acceleration observation series to build a hidden Markov model (HMM)to detect the fall behavior. The method extracted acceleration characteristic time series from human fall course to describe the fall process, and used the acceleration characteristic time series to train HMM in order to build a random process mathematical model. The 300 samples of experimental data from 10 volunteers were obtained, and 5-fold cross-validation was used to estimate the model. Results showed that the accuracy of the method was 98.2%, the sensitivity was 91.3%, and the specificity was 99.6%, showing that the proposed method gets good result in detecting fall events.
曹荟强, 林仲志, 吴水才. 基于隐马尔可夫模型的老年人跌倒行为检测方法研究[J]. 中国生物医学工程学报, 2017, 36(2): 165-171.
Cao Huiqiang, Lin Chungchih, Wu Shuicai. A Detection Method for the Fall Behavior of Elders Based on Hidden Markov Model. Chinese Journal of Biomedical Engineering, 2017, 36(2): 165-171.
[1] 朱月妹, 袁浩斌, 陈雷. 老年人跌倒危险因素的调查 [J]. 护理实践与研究, 2007, 4(10): 5-7. [2] 郝燕萍, 刘雪琴, 靳海如. 老年人跌倒伤残情况分析 [J]. 护理研究, 2006, 20(8): 2017-2018. [3] Rougier C, Meunier J. Demo: Fall detection using 3D head trajectory extracted from a single camera video sequence [J]. Journal of Telemedicine and Telecare, 2005, 11(4): 37-42. [4] Shoaib M, Dragon R, Ostermann J. View-invariant fall detection for elderly in real home environment [C] //2010 Fourth Pacific-Rim Symposium on Image and Video Technology. Singapore: IEEE, 2010: 52-57. [5] Qian Huimin, Mao Yaobin, Xiang Wenbo, et al. Recognition of human activities using SVM multi-class classifier [J]. Pattern Recognition Letters, 2010, 31(2): 100-111. [6] Khan Z A, Sohn W. Abnormal human activity recognition system based on R-transform and kernel discriminant technique for elderly home care [J]. IEEE Transactions on Consumer Electronics, 2011, 57(4): 1843-1850. [7] Nyan MN, Tay FE, Murugasu E. A wearable system for pre-impact fall detection [J]. Journal of Biomechanics, 2008, 41(16):3475-3481. [8] Li Q, Stankovic JA, Hanson MA, et al. Accurate, fast fall detection using gyroscopes and accelerometer-derived posture information[C]//2009 6th International Workshop on Wearable and Implantable Body Sensor Networks. Berkeley: IEEE, 2009:138 - 143. [9] Tong Lina, Chen Wei, Song Quanjun, et al. A research on automatic human fall detection method based on wearable inertial force information acquisition system [C] //2009 IEEE International Conference on Robotics and Biomimetics. Bangkok: IEEE, 2009: 949-953. [10] Chen GC, Huang CN, Chiang CY, et al. A reliable fall detection system based on wearable sensor and signal magnitude area for elderly residents [C] //International Conference on Smart Homes and Health Telematics. Seoul: Springer, 2010: 267-270. [11] Goh ATC. Back-propagation neural networks for modeling complex systems [J]. Artificial Intelligence in Engineering, 1995, 9(3): 143-151. [12] 王之琼, 曲璐渲, 隋雨彤, 等. 基于极限学习机的跌倒检测分类识别研究 [J]. 中国生物医学工程学报, 2014, 33(4): 418-424. [13] Li Yanjun, Chen Gan, Shen Yueyun, et al. Accelerometer-based fall detection sensor system for the elderly [C] //2012 2nd International Conference on Cloud Computing and Intelligence Systems. Hangzhou: IEEE, 2012, 3: 1216-1220. [14] Sorvala A, Alasaarela E, Sorvoja H, et al. A two-threshold fall detection algorithm for reducing false alarms [C] //2012 6th International Symposium on Medical Information and Communication Technology. La Jolla: IEEE, 2012: 1-4. [15] Dumitrache M, Pasca S. Fall detection algorithm based on triaxial accelerometer data [C] //2013 E-Health and Bioengineering Conference. Iasi: IEEE, 2013: 1-4. [16] Bourke AK, O’brien JV, Lyons GM. Evaluation of a threshold-based tri-axial accelerometer fall detection algorithm [J]. Gait & posture, 2007, 26(2): 194-199. [17] Yamato J, Ohya J, Ishii K. Recognizing human action in time-sequential images using hidden markov model [C] //1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Champaign: IEEE, 1992: 379-385. [18] Lee KF, Hon HW. Speaker-independent phone recognition using hidden Markov models [J]. Acoustics, Speech and Signal Processing, IEEE Transactions on, 1989, 37(11): 1641-1648. [19] Nathasitsophon Y, Auephanwiriyakul S, Theera-Umpon N. Fall detection algorithm using linear prediction model [C] //2013 IEEE International Symposium on Industrial Electronics. Taipei: IEEE, 2013: 1-6. [20] Tong Lina, Song Quanjun, Ge Yunjian, et al. HMM-based human fall detection and prediction method using tri-axial accelerometer [J]. IEEE Sensors Journal, 2013, 13(5): 1849-1856. [21] He Jin, Li Huanming, Tan Jindong. Real-time daily activity classification with wireless sensor networks using hidden markov model [C] //29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Lyon: IEEE, 2007: 3192-3195. [22] 卢先领, 王洪斌, 王莹莹, 等. 一种基于加速度传感器的人体跌倒识别方法 [J]. 计算机应用研究, 2013, 30(4): 1109-1111. [23] 佟丽娜, 宋全军, 葛运建. 基于时序分析的人体摔倒预测方法 [J]. 模式识别与人工智能, 2012, 25(2): 273-279. [24] Rabiner LR, Juang BH. An introduction to hidden Markov models [J]. IEEE ASSP Magazine, 1986, 3(1): 4-16.
