使用智慧型手機實現人類動作狀態識別

Implementation of the Human Action State Recognition Using Smartphone

陳冠宇(Kuan-Yu Chen)

指導教授 ： 廖俊鑑

朝陽科技大學/資訊學院/資訊與通訊系/碩士(2016年)

摘要

依據104年「運動城市調查」報告發現近年來越來越多台灣民眾從事慢跑、籃球、腳踏車、爬山和游泳等運動，而從事走路和慢跑的運動為70.2%的比例，所以本研究針對走路與跑步運動進行辨識。
本研究使用智慧型手機內加速感測器辨識人類動作狀態。進而使用九種特徵值分別為訊號強度面積(Signal Magnitude Area)、變異值(Variation)、相關度(Correlation)、峰度(Kurtosis)、偏度(Skewness)、振幅(Amplitude)、差分訊號(Differential Signal Magnitude)、平均值(Mean)和標準差(Standard Deviation)辨識動作狀態。由實驗結果上述九種特徵值中有四種特徵值能使用辨識靜止動作狀態，走路動作狀態有四種特徵值能辨識，六種特徵值能辨識跑步動作狀態。進而實驗五名受測者進行獨立跑步、走路與靜止動作和連續跑步、走路與靜止動作特徵值辨識，具有良好的動作辨識正確率。

關鍵字

特徵值 ； 三軸加速感測器 ； 動作辨識

並列摘要

Based on "Sports City Survey" report, 2015, it was found that in recent years more and more people in Taiwan are engaging in jogging, basketball, cycling, climbing and swimming and other sports; and 70.2 % of sport people engage in walking and jogging motion, so this study is focused on walking and running identification.
This study uses acceleration sensors provided by smartphones to identify human action state. Furthermore, identification of the operation status is accomplished by nine kinds of feature values: Signal Magnitude Area, Variation, Correlation, Kurtosis, Skewness, Amplitude, Differential Signal Magnitude, Mean and Standard Deviation. From the experimental results, only four characteristic values in the nine feature can be used to identify a stationary operating state; four feature can identify walking state; six feature can identify running motion state. Further, five persons are subject to identification experiments such as independent running, walking and stationary action and continuous running, walking and stationary action, which are all with good action recognition accuracy.

並列關鍵字

Feature ； Accelerometer ； Action identified

參考文獻

[2]A. Bayat, M. Pomplun and D. A. Tran, “A Study on Human Activity Recognition Using Accelerometer Data from Smartphones,” The 11th International Conference on Mobile Systems and Pervasive Computing, Vol. 46, pp. 540-547, 2014.

[5]S. Park and S. Jayaraman, “Enhancing the Quality of Life Through Wearable Technology,” Engineering in Medicine and Biology Magazine, IEEE, Vol. 22, No. 3, pp. 41-48, 2003.

[6]T. Choudhury, S. Consolvo, B. Harrison, and J. Hightower, “The Mobile Sensing Platform: An Embedded Activity Recognition System,” IEEE Pervasive Computing, Vol. 7, No. 2, pp. 32-41, 2008.

[8]T. P. Kao, C. W. Lin and J. S. Wang, “Development of a Portable Activity Detector for Daily Activity Recognition,” IEEE International Symposium on Industrial Electronics (ISIE 2009), pp. 5-8, 2009.

[9]V. Lugade, E. Fortune, M. Morrow, K. Kaufman, “Validity of using tri-axial accelerometers to measure human movement-Part I: Posture and movement detection,” Medical Engineering & Physics, Vol. 36, Issue 2, pp.169-176, 2014.