本論文主旨在於以三軸加速度計之訊號進行行走模式分類與距離估測演算法的開發。在實驗訊號收集上，使用者配戴三軸加速度計於慣用腳之腳踝來收集日常生活常見之行走模式加速度訊號(平地行走、上樓及下樓)及相對應的行走距離。在行走模式分類以及距離估測演算法，本論文以行走中步態週期的細部相位資訊做為開發行走模式分類以及距離估測演算法的基礎。本論文首先將收集到的加速度訊號進行三軸合力計算與視窗化等前處理程序，接著基於前處理過後之加速度合力訊號，搭配站立期偵測及動態閥值偵測，來完成步態週期分析中站立期、推蹬期、擺動期與踏地期之相位區間偵測，然後基於步態相位偵測結果作為基礎，作為行走模式分類與距離估測演算法所需之基本資訊。
在行走模式分類方面，總共有20位受測者(14男6女)進行實驗收案，其實驗動作主要包含了平地行走、上樓以及下樓動作，以自我主觀之正常速度進行加速度訊號的收集，再依各動作訊號透過步態相位偵測所獲得之步態相位之時間特性，搭配決策樹的建構以完成行走模式分類演算法；另一方面，在距離估測演算發開發方面，總共有9位受測者(7男2女)進行實驗收案，並以平地行走為實驗動作。在進行距離估測演算法上，可分為步數估測及步長估測兩部份。首先在步數估測方面，本論文以步態相位偵測作為基礎，透過步態相位循環特性來定義步數數目；而在步長估測上，本文以步長、步速、以及步頻等參數共同建立迴歸方程式以估測距離。其中步長的參數以受測者個人基本資料(身高)、以及 速度、步頻來進行步長迴歸方程式的建置；接著，量測之步態加速度訊號必頇以步態相位偵測結果作為基礎，再取用步態週期之暫態加速度訊號來擷取腳踝加速度訊號之單軸、雙軸及三軸之向量合力的變異數、平均數以及中位數來完成速度迴歸方程式。
本論文開發了結合行走模式分類與距離估測演算法來幫助使用者了解自我身體每天的步行活動狀況，且行走模式分類演算法對於日常生活常見之帄地行走、上樓及下樓動作其辨識率可達96.44%，而於距離估測與步數估測則分別具有96.42%和99.75%的平均正確率。

This thesis presents a series of algorithms for walking pattern classification and distance estimation using a triaxial accelerometer. The triaxial acceleration signal is collected from the accelerometer module worn on users’ ankle, which is used for analyzing walking patterns (level walking, upstairs, and downstairs) and its corresponding walking distance. In order effectively analyzing walking patterns and distance, we propose a gait phase analysis algorithm including a stance phase detection and a dynamic threshold detection algorithm to discriminate gait phases including stance, push off, swing, and heel strike phase. Based on the information of gait phases, we further develope algorithms for walking patterns classification and distance estimation.
In order to validate the effectiveness of using accelerometer in the walking pattern classification, an indoor experiment with 20 subjects (6 females and 14 males) including level walking, upstairs, and downstairs was conducted. The gait phase information was extracted to construct a decision tree for identifying walking patterns. In the walking distance estimation, the outdoor experiment including level walking was performed on 9 subjects (2 females and 7 males) with free speeds. The walking distance estimation algorithm consists of walking steps estimation and step length estimation. The walking steps are obtained by the gait phase detection, and the step length estimation is based on the information of height of subjects, step speed, and step frequency. Finally, the walking distance estimation can be obtained.
Walking pattern classification and distance estimation algorithms have been successfully developed to assist users to understand their walking activities information. In our experiments, the average recognition rate of walking activities using the proposed decision tree is 96.44%, and the average accuracy rates of walking distance estimation with step length regression model and step counting are 96.42% and 99.75%.

[1] Y. Schutz, S. Weinsier, P. Terrier, and D. Durrer, “A new accelerometric method to assess the daily walking practice,” International Journal of Obesity, vol. 26, no. 1, pp. 111-118, 2002.
[2] L. E. Bratteby, B. Sandhagen, M. Lotborn, and G. Samuelson, “Daily energy expenditure and physical activity assessed by an activity diary in 374 randomly selected 15-year-old adolescents,” Eur. J. Chin. Nutr., vol. 51, pp. 592-600, 1997.
[3] P. Barralon, N. Noury, and N. Vuillerme, “Classification of daily physical activities from a single kinematic sensor,” Proc. of the 2005 IEEE 27th Annual Conference on Engineering in Medicine and Biology, 2005, pp. 2447-2450.
[4] M. Sekine, T. Tamura, T. Togawa, and Y. Fukui, “Classification of waist-acceleration signals in a continuous walking record,” Med. Eng. Phys., vol. 22, no. 4, pp. 285-291, 2000.
[5] S. Godha and G. Lachapelle, “Foot mounted inertial system for pedestrian navigation,” Meas. Sci. Technol., vol. 19, no. 7, pp. 075202-075210, 2008.
[6] Z. Sun, X. Mao, W. Tian, and X. Zhang, “Activity classification and dead reckoning for pedestrian navigation with wearable sensors,” Meas. Sci. Technol., vol. 20, no. 1, pp. 015203-015212, 2009.
[7] C. Huang, Z. Liao, and L. Zhao, “Synergism of INS and PDR in self-contained pedestrian tracking with a miniature sensor module,” IEEE Sensors J., vol. 10, no. 8, pp. 1349-1359, 2010.
[8] E. Foxlin, “Pedestrian tracking with shoe-mounted inertial sensors,” IEEE Comput. Graphics Appl., vol. 25, no. 6, pp. 38-46, 2005.
[9] J. A. Lee, S. H. Cho, Y. J. Lee, H. K. Yang, and J. W. Lee, “Portable activity monitoring system for temporal parameters of gait cycles,” J. Med. Syst., vol. 34, no. 5, pp. 959-966, 2009.
[10] S. J. Bamberg, A. Y. Benbasat, D. M. Scarborough, D. E. Krebs, and J. A. Paradiso, “Gait analysis using a shoe-integrated wireless sensor system,” IEEE Trans. Inf. Technol. Biomed., vol. 12, no. 4, pp. 413-423, 2008.
[11] B. Coley, B. Najafi, A. Paraschiv-Ionescu, and K. Aminian, “Stair climbing detection during daily physical activity using a miniature gyroscope,” Gait and Posture, vol. 22, no. 4, pp. 287-294, 2005.
[12] M. Windolf, N. Götzen, and M. Morlock, “Systematic accuracy and precision analysis of video motion capturing systems-exemplified on the Vicon-460 system,” J. Biomech., vol. 41, no. 12, pp. 2776-2780, 2008.
[13] 行政院衛生署國民健康局，每日行走建議活動量及其說明，民國91年。
[14] 行政院衛生署國民健康局，國民運動行為調查結果說明會，民國94年。
[15] C. E. Baudy, and A. D. Kuo, “Active control of lateral balance in human walking,” J. Biomech., vol. 33, no. 11, pp. 1433-1440, 2000.
[16] A. TH. M. Willemsen, F. Bloemhof, and H. B. K. Boom, “Automatic stance-swing phase detection from accelerometer data for peroneal nerve stimulation,” IEEE Trans. Biomed. Eng., vol. 37, no. 12, pp.1201-1208, 1990.
[17] I. P. I. Pappas, M. R. Popovic, T. Keller, V. Dietz, and M. Morari, “A reliable gait phase detection system,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 9, no. 2, pp. 113-125, 2001.
[18] T. Liu, Y. Inoue, K. Shibata, and X. Tang, “A wearable inertial sensor system for human motion analysis,” Proc. IEEE Int. Symp. Comput. Intell. Robot. Autom. CIRA, 2005, pp.409-413.
[19] D. Kotiadis, H. J. Hermens, and P. H. Veltink, “Inertial Gait Phase Detection for control of a drop foot stimulator. Inertial sensing for gait phase detection,” Med. Eng. Phys., vol. 32, no. 4, pp. 287-297, 2010.
[20] R. C. González, A. M. López, J. Rodriguez-Uría, D. Álvarez, and J. C. Alvarez, “Real-time gait event detection for normal subjects from lower trunk accelerations,” Gait and Posture, vol. 31, no. 3, pp. 322-325, 2010.
[21] T. Tamura, M. Sekine, M. Ogawa, T. Togawa, and Y. Fukii, “Classification of acceleration waveforms during walking by wavelet transform,” Methods Inform. Med., vol. 36, pp. 356-359, 1997.
[22] J. Staudenmayer, D. Pober, S. Crouter, D. Bassett, and P. Freedson, “An artificial neural network to estimate physical activity energy expenditure and identify physical activity type from an accelerometer,” Journal of Applied Physiology, vol. 107, no. 4, pp. 1300-1307, 2009.
[23] M. Sekine, T. Tamura, M. Akay, T. Fujimoto, T. Togawa, and Y. Fukui, “Discrimination of walking patterns using wavelet-based fractal analysis,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 10, no. 3, pp. 188-196, 2002.
[24] A. Salarian, H. Russmann, F. J. G. Vinqerhoets, C. Dehollain, Y. Blanc, P. R. Buckhard, and K. Aminian, “Gait assessment in Parkinson's disease: Toward an ambulatory system for long-term monitoring,” IEEE Trans. Biomed. Eng., vol. 51, no. 8, pp. 1434-1443, 2004.
[25] M. Bächlin, M. Plotnik, D. Roggen, I. Maidan, J. M. Hausdorff, N. Giladi, and G. Tröster, “Wearable assistant for Parkinsons disease patients with the freezing of gait symptom,” IEEE Trans. Inf. Technol. Biomed., vol. 14, no. 2, pp. 436-446, 2010.
[26] N. L. Dudek, O. D. Khan, E. D. Lemaire, M. B. Marks, and L. Saville, “Ambulation monitoring of transtibial amputation subjects with patient activity monitor versus pedometer,” J. Rehabil. Res. Dev., vol. 45, no. 4, pp. 577-586, 2008.
[27] M. G. Benedetti, F. Catani, T. W. Bilotta, M. Marcacci, and S. Giannini, “Muscle activation pattern and gait biomechanics after total knee replacement,” Clin. Biomech., vol. 18, no. 9, pp. 871-876, 2003.
[28] B. J. Jaeger, S. A. Olivares, and M. C. Wetzel, “Operant conditioning in relation to natural EMG during rapid human walking,” Am. J. Phys. Med., vol. 66, no. 2, pp. 59-76, 1987.
[29] 戴久永，統計概念與方法，民國88年8月。
[30] S. B. Provost, Y. H. Cheong, “On approximating the distribution of the Durbin-Watson statistic from its moments obtained recursively,” Am. J. Math. Manage. Sci., vol. 25, no. 3-4, pp. 207-220, 2005.
[31] S. Studenski, S. Perera, K. Patel, C. Rosano, K. Faulkner, M. Inzitari, J. Brach, J. Chandler, P. Cawthon, E. B. Connor, M. Nevitt, M. Visser, S. Kritchevsky, S. Badinelli, T. Harris, A. B. Newman, J. Cauley, L. Ferrucci, and J. Guralnik. “Gait Speed and survival in older adults,” JAMA., vol. 305, no. 1, pp. 50-58, 2011.