忙碌的現代人為了保持健康越來越重視運動，以台灣為例，運動人口比例也逐年增長，且由於疫情影響，居家健身風氣盛行，啞鈴便成為熱門運動器材。此外，運動產業的風貌也隨科技的不斷進步而改變，智慧化的運動訓練已逐漸成為趨勢，而運動活動的自動紀錄和分析可以幫助使用者監控自己的運動狀況並發揮最大程度的訓練效果。
為了有效紀錄運動成果，本研究將具慣性感測器（IMU）的動作感測裝置安裝於啞鈴上，用以擷取運動時的加速度與陀螺儀訊號，並開發一套應用於啞鈴運動分析的演算法，可以自動監測運動時的動作及次數，共包含六種常見的啞鈴訓練動作。演算法分為兩部分：第一部分為訊號預處理，先將從慣性感測器收集到之原始加速度及陀螺儀訊號進行處理，排除多餘雜訊；第二部分為訊號分析，從分割好之訊號中計算特徵值，再以皮爾遜積差相關係數（Pearson's）做排序，最後將經特徵選取後的特徵子集套入機器學習，並以KNN（K-nearest-neighbor）作為動作分類器。
為了評估即時的動作辨識能力，將演算法在Android Application中實現，實驗結果顯示，兩手的平均運動次數及動作辨識準確度分別可以達到為99.6%及98.8%，結果表明，該方法是可行的。

In modern society, people pay more and more attention on exercise to maintain a healthy body. With the rise of health awareness, Taiwan’s sport population has increased annually. Because home fitness is extremely prevalent during the coronavirus pandemic, the population of using dumbbells is increased. In addition, the development of the sports industry has also changed with the advance in technology. Intelligent fitness equipment is becoming a trend. It aims to help athletes to monitor their progress and to maximize their training effects.
In order to effectively record exercise results, we use a dumbbell equipped with the inertial measurement unit containing an accelerometer and a gyroscope. An algorithm is developed to analyze the information during exercises. This work is designed to recognize six common dumbbell training movements and count repetitions. The algorithm is divided into two parts. The first part is data pre-processing, which mainly eliminates the noise. The second part is data analysis. The feature values are calculated from the split signals, and then a statistical measure (Pearson's) is applied to assign a scoring to each feature. The features are ranked by the scores and selected to compile a feature subset. In the classification, this thesis uses KNN （K-nearest neighbor） as the classification of activities.
To evaluate the real time capability, the algorithm was run on an Android application. The experimental results show that the algorithm of this study can be adapted to both hands. The mean accuracy of 99.6% and 98.8% is achieved for activity recognition and repetition count, respectively. The performance demonstrates the feasibility of the proposed approach.

[1] i運動資訊平台-教育部體育署, “中華民國108年運動現況調查結案報告書,” 2019
[2] Centers for Disease Control and Prevention, “How much physical activity do adults need?” Accessed: Jun. 26, 2020. [Online]. Available: https://www.cdc.gov/physicalactivity/basics/adults/index.htm
[3] Adobe Analytics, “Adobe Unveils First Digital Economy Index,” Accessed: Jun. 26, 2020. [Online]. Available: https://s23.q4cdn.com/979560357/files/033120AdobeDigitalEconomyIndex.pdf
[4] I. Ar, and Y. S. Akgul, “A Computerized Recognition System for the Home-Based Physiotherapy Exercises Using an RGBD Camera,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 22, no. 6, pp. 1160-1171, May 2014.
[5] Q. Nie, J. Wang, X. Wang, and Y. Liu, "View-Invariant Human Action Recognition Based on a 3D Bio-Constrained Skeleton Model," IEEE Transactions on Image Processing, vol. 28, no. 8, pp. 3959-3972, Aug. 2019.
[6] C. Li, M. Fei, H. Hu, and Z. Qi, “Free Weight Exercises Recognition Based on Dynamic Time Warping of Acceleration Data,” in Proc. Communications in Computer and Information Science, 2013, pp. 178-185.
[7] B. J. Mortazavi, M. Pourhomayoun, G. Alsheikh, N. Alshurafa, S. I. Lee, and M. Sarrafzadeh, “Determining the Single Best Axis for Exercise Repetition Recognition and Counting on SmartWatches,” in Proc. 2014 11th International Conference on Wearable and Implantable Body Sensor Networks, 2014, pp. 33-38.
[8] P. Hausberger, A. Fernbach, and W. Kastner, “IMU-based Smart Fitness Devices for Weight Training,” in Proc. IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society, 2016, pp. 5182-5189.
[9] D. Das, S. M. Busetty, V. Bharti, and P. K. Hegde, “Strength Training: A Fitness Application for Indoor Based Exercise Recognition and Comfort Analysis,” in Proc. 2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA), 2017, pp.1123-1129.
[10] R. Ma, D. Yan, H. Peng, T. Yang, X. Sha, Y. Zhao, and L. Liu, “Basketball Movements Recognition Using a Wrist Wearable Inertial Measurement Unit,” in Proc. 2018 IEEE 1st International Conference on Micro/Nano Sensors for AI, Healthcare, and Robotics (NSENS), 2018, pp. 73-76.
[11] A. L. S. V. Siddhartha Varma, V. Kumar Sinha, and D. Das, “Health Assessment for Office Workers by Tracking the Wrist Motion Centric Activity,” in Proc. 2019 IEEE 16th India Council International Conference (INDICON), 2019, pp. 1-4.
[12] R. Mahony, T. Hamel and J. Pflimlin, "Nonlinear Complementary Filters on the Special Orthogonal Group," IEEE Transactions on Automatic Control, vol. 53, no. 5, pp. 1203-1218, Jun. 2008.
[13] M. Euston, P. Coote, R. Mahony, J. Kim, and T. Hamel, “A Complementary Filter for Attitude Estimation of a Fixed-Wing UAV,” in 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2008, pp. 340-345.
[14] W. R. Hamilton, “On Quaternions; or on a new System of Imaginaries in Algebra,” London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, vol. 25, pp. 489-495, Jul. 1844.
[15] J. B. Kuipers, “Quaternions and Rotation Sequences,” in Proc. Geometry, Integrability and Quantization, Sep. 1-10, 1999, pp. 127-143.
[16] MathWorld, “Rodrigues' Rotation Formula,” Accessed: Jun. 25, 2020. [Online]. Available: https://mathworld.wolfram.com/RodriguesRotationFormula.html
[17] V. Bachu and J. Anuradhu, “A Review of Feature Selection and Its Methods,” Cybernetics and Information Technologies, vol. 19, pp. 3-26, Mar. 2019.
[18] H. Liu, M. Zhou, and Q. G. Liu, “An Embedded Feature Selection Method for Imbalanced Data Classification,” IEEE/CAA Journal of Automatica Sinica, vol. 6, no. 3, pp. 703-715, May 2019.
[19] J. C. Platt, “Sequential Minimal Optimization: A Fast Algorithm for Training Support Vector Machines,” Advances in Kernel Methods-Support Vector Learning, vol. 208, Jul. 1998.
[20] L. Breiman, “Random Forests,” Machine Learning, vol. 45, pp. 5-32, Oct. 2001.
[21] A. McCallum, and K. Nigam, “A Comparison of Event Models for Naive Bayes Text Classification”, Work Learn Text Categ, vol. 752, May 2001.
[22] N. S. Altman, “An Introduction to Kernel and Nearest-Neighbor Nonparametric Regression,” American Statistician, vol. 46, no. 3, pp. 175-185, Aug. 1992.
[23] Nordic Semiconductor, “nRF52832,” Accessed: Jun. 23, 2020. [Online]. Available: https://www.nordicsemi.com/Products/Low-power-short-range-wireless/nRF52832
[24] InvenSense, “ICM-20649,” Accessed: Jun. 23, 2020. [Online]. Available: https://product.tdk.com/info/en/documents/catalog_datasheet/imu/DS-000192-ICM-20649-v1.0.pdf
[25] Dumbbell-Exercises.Com, " Dumbbell Workout Routines," Accessed: Mar. 12, 2020. [Online]. Available: https://dumbbell-exercises.com/workout-routines/#5
[26] J. Diebel, “Representing Attitude: Euler Angles, Unit Quaternions, and Rotation Vectors,” Matrix, vol. 58, no. 15–16, pp. 1-35, Jan. 2006.
[27] K. P. Lin, W. H. Chang, “A Technique for Automated Arrythmia Detection of Holter ECG,” in Proc. 17th International Conference of the Engineering in Medicine and Biology Society, Sept. 1995, vol 1, pp. 183-184
[28] U. Jensen, M. Ring, and B. Eskofier, “Generic Features for Biosignal Classification,” Sportinformatik 2012, vol. 112, pp. 162–168, 2012.
[29] Package weka.attributeSelection, ” Class CorrelationAttributeEval,” Accessed: May 13, 2020. [Online]. Available: http://infochim.u-strasbg.fr/cgi-bin/weka-3-9-1/doc/index.html?weka/attributeSelection/CorrelationAttributeEval.html