透過您的圖書館登入
IP:3.141.199.243
  • 學位論文

利用動態波形偵測和雙閥值技術實現一個以加速規為基礎的游泳分析系統

Implementing An Accelerometer-Based Swimming Analysis System Using Dynamic Waveform Detection and Dual Thresholding Techniques

指導教授 : 繆紹綱
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


隨著生活水準的提升,人們對自身健康的問題越來越重視,而想要獲得健康,適當的運動不可少。本研究選擇游泳運動為探討對象是因為該運動為全身性運動,對人體健康有極大的助益,且幾乎沒有年齡上的限制。游泳時,若能自動記錄游泳揮臂次數以及游泳來回趟數等資訊,對休閒型或訓練型的游泳者來說都更加便利。 本研究將探討如何利用三軸加速度感應器來估計游泳的揮臂次數、來回泳池的趟數、每趟所花的時間與速度。本研究提出的方法主要是(1)揮臂波形動態偵測法:此法可以有效將濾除掉雜訊後的訊號再進一步修正,並運算出加速度波形中每次揮臂的波峰與波谷位置,再判斷該波峰與波谷是否為游泳開始的位置;(2)雙閥值游泳分析法:主要是透過高低兩個不同閥值來掃描加速度波形,並判斷是否正在游泳與計算游泳的揮臂次數。 本研究的做法與利用拍攝影像的傳統方式相比,可以提供(1)水下游泳資訊(傳統方法須架設水下攝影機) ;(2)可更便利提供與游泳相關的即時資訊。本研究的系統實際與業界商品做比較,在總體揮臂錯誤率上,本研究的系統為2.05%而業界商品為4.51%,由此可以知道本研究所提系統的準確度已達到目前業界的水準甚至超越它。由於本研究是在已知游泳姿勢下所做的探討,未來可以結合游泳姿勢自動偵測,提供更多相關資訊的回饋並建立雲端系統做游泳活動的長期記錄與分析。

並列摘要


As the quality of life has been improved, people are more and more concerned about their own health. People who want to keep their health should do the exercise regularly. This thesis chooses swimming to be the research topic because it involves whole body movement which is good for health and there is no age limit to do this exercise. While swimming, it will be convenient if professional or casual swimmers can record their stroke and lap information automatically. How to apply a three-axis acceleration sensor in order to evaluate stroke frequency, number of laps, time required per lap and speed per lap will be discussed in this study. The proposed system consists of two major parts. First, a dynamic stroke waveform detection algorithm: it is effective to fix a noise-reduced and filtered signal further, identity the peak and the valley points of the acceleration waveform for stroke detection, and use the peak and valley points to determine when the swimming movement is started. Second, a dual-threshold swimming analysis algorithm: it uses high and low thresholds to scan acceleration waveform and determine whether a swimmer is actually swimming and count the number of swimming stroke. As compared to the traditional method of taking video, the proposed system can provide (1) underwater swimming information (the traditional method needs an underwater camera in this case); (2) real-time swimming information more conveniently. The error counts of total strokes obtained with the proposed method in this thesis and a commercial product are 2.18% and 4.51%, respectively. This result shows that the accuracy of the proposed method is better than the commercial product. The study here is based on the assumption that the swimming type is known. Thus, the future works of this study may include the automatic detection of swimming types, more relevant information feedback from swimming type detection and the use of cloud computing for long-tern recording and analysis of swimming activity.

參考文獻


[29] 張粳魁,利用卡曼濾波器整合全球定位系統及慣性導航系統,國立台北科技大學機電整合研究所碩士論文,2008。
[13] 邱文信,擺臂式出發即接力出發技術分析,育達研究叢刊,Vol. 3, pp. 163-174, 2004。
[2] 洪連進、韓大衛和盧居福,「台北市運動中心羽球、健身、舞蹈運動休費者休閒運動效益之研究」,臺北市立教育大學學報,Vol. 4,No. 1,pp. 27-48,2010。
[11] 王甯和劉有德,從動作協調探討提升游泳速度之要素,中華體育,Vol. 21, No. 3, pp. 135-142, 2007。
[1] 黃蕙娟,健身俱樂部相關法規之初探,中華體育季刊,Vol. 17, No. 4, pp. 153-162, 2003。

被引用紀錄


張倩瑜(2014)。一個以三軸加速規訊號為基礎的自動化游泳姿勢辨識系統〔碩士論文,中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu201401008

延伸閱讀