隨著智慧型手機的發展與固態微機電(MEMS)的普及與其技術發展，在手機內嵌微機電系統，或稱慣性感測器讓智慧型手機有著多元化的發展可能性。本論文嘗試使用Android智慧型手機平台實作結合行為辨識的智慧型計步器，讓我們除了系統計步器功能並可以同時計算和取得使用者的行為。在計步器方面使用動態測量的概念，利用雙層平均化濾波的方法達到去除雜訊與同時取得動態平均的效果，令步伐測量的準確度不會依賴手機的指向方位。並將計步器結合行為辨識，利用即時的更新使用者行為讓手機盡量不利用閥值偵測步伐同時提供更多使用者資訊給予參考。實驗結果顯示在我們的即時行為辨識系統之下，擁有的動態回應時間在系統實驗之下均少於5秒。而步伐偵測和行為辨識的精準度與其參考的研究論文比較更為精確。

As the development of the smart phone and MEMS technology comes to more normalization, put inertial sensor into smart is trend and it leads to high range develop possibilities of smart phone. In our thesis, we discussed the implement of pedometer in android platform with activity recognition. User not only get the number of step user takes, but also have information with user’s activity in the same time. In part of pedometer, we use the method of 2-layer moving average filter to remove signal noise and dynamic average. The pedometer’s accuracy should not depend critically on the relationship between the motion axes and the accelerometer’s measurement axes. Meanwhile, we integrate real-time activity recognition method and enhance filter of signal processing, lead to better performance and user information by remove decision of threshold value. In the experiment result, our design of real time activity recognition system depicts that we have dynamic response time which is less than 5 second. As to performance of pedometer and activity recognition, we compare our system with reference paper and obtain better accuracy.

Chapter 1 Introduction
Chapter 2 Related Work
2.1 Step Detection
2.2 Activity Recognition
Chapter 3 System Implement and Design
3.1 System Prototype
3.2 Step Count Algorithm
3.2.1 Signal filter by sliding window
3.2.2 Cyclic Pattern Recognition
3.3 Activity Recognition Algorithm
3.3.1 Feature Extraction
3.3.2 K-Nearest Neighbor algorithm
3.4 System Design and approach
3.4.1 Flow chart of system design
3.4.2 Real-time activity recognition mechanism
3.4.3 Performance enhance mechanism
Chapter 4 Experiment and Evaluation
4.1 Accuracy of step count
4.2 Accuracy of activity recognition
Chapter 5 Conclusion
References

[1] P. Alto, “Google’s Android becomes the world’s leading smart phone platform,” Canalys research release, Jan. 2011.
[2] J. Wang, “China smart phone market 2012: Trends and analysis,” DIGITIMES Research, Taipei, July, 2012.
[3] Android Developer Challenge. https://developers.google.com/
[4] S. I. Yang and S. B. Cho, “Recognizing Human Activities from Accelerometer and Physiological Sensors,” Proceeding of IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, pp. 100-105, Seoul, Aug. 2008.
[5] S. Y. Cho and C. G. Park, “MEMS Based Pedestrian Navigation System,” International Journal of Navigation, Vol. 10, No. 59, pp. 135-153, 2006.
[6] R. Libby, “A simple method for reliable footstep detection on embedded sensor platforms,” http://ubicomp.cs.washington.edu/uwar/libby_peak_detection.pdf, Accessed December, 2009.
[7] M. Romanovas and M. Schwaab, “A Study on Indoor Pedestrian Localization Algorithms with Foot-Mounted Sensors,” Proceeding of 13th-15th International Conference on Indoor Positioning and Indoor Navigation, pp. 1-10, Sydney, Australia, Date 13-15 Nov. 2012.
[8] T. Chen, W. Hu and R. Sun, “Displacement Measurement Algorithm Using Handheld Device with Accelerometer,” Proceeding of Asia-Pacific Conference on Wearable Computing Systems, pp. 122-126, Shenzhen, China, April 2010.
[9] S. Colton, “The Balance Filter, A Simple Solution for Integrating Accelerometer and Gyroscope Measurements for a Balancing Platform,” Chief Delphi white paper, http://web.mit.edu/scolton/www/filter.pdf, 2007.
[10] G. Trehard and M. Boukallel, ”Indoor Pedestrian Localisation Solution based on Anemometry Sensor Integration with a Smartphone,” Proceeding of 13th-15th International Conference on Indoor Positioning and Indoor Navigation, pp. 1-9, Sydney, Australia, Date 13-15 Nov. 2012.
[11] F. Keller and T. Willemsen, “Calibration of Smartphones for the use in indoor navigation,” Proceeding of 13th-15th International Conference on Indoor Positioning and Indoor Navigation, pp. 84-92, Sydney, Australia, Date 13-15 Nov. 2012.
[12] R. Jirawimut and P. Ptasinski, ”A Method for Dead Reckoning Parameter Correction in Pedestrian Navigation System” Proceeding of IEEE Transactions on Instrumentation and Measurement, Vol. 52, No. 1, pp. 209-215, Feb. 2003.
[13] S. Jayalath, J. Syrjarinne, and N. Abhayasinghe, “A gyroscopic data based pedometer algorithm,” Proceeding of the 8th International Conference Computer Science & Education (ICCSE), pp. 1369-1373, Colombo, Sri Lanka, April 2013.
[14] G. S. Kolt, G. M. Schofield, N. Kerse, N. Garrett, and T. Ashton, “Healthy Steps Trial: Pedometer-Based Advice and Physical Activity for Low-Active Older Adults,” Annals of Family Medicine in the U.S. National Institutes of Health, Vol. 10, No. 3, May 2012.
[15] F. Scholkmann, J. Boss and M. Wolf, “An Efficient Algorithm for Automatic Peak Detection in Noisy Periodic and Quasi-Periodic Signals,” Algorithm, Vol. 5, No. 4, pp. 588-603, 2012.
[16] M. Martin and M. Michael, “A Step Counter Service for Java-Enabled Devices Using a Built-In Accelerometer,” Proceeding of the International Conference on Communication System Software and Middleware, pp. 1-5, Dublin, Ireland, April 2009.
[17] N. Zhao, “Full-Featured Pedometer Design Realized with 3-Axis Digital Accelerometer,” Analog Dialogue, pp 44-50, 2010.
[18] T. Schlomer, B. Poppinga and N. Hemze, “Gesture recognition with a Wii controller,” Proceeding of the 2nd international conference on Tangible and embedded interaction, pp. 11-14, New York, Feb. 2008.
[19] X. Long, B. Yin and R. M. Aarts, “Single-Accelerometer-Based Daily Physical Activity Classification,” Proceeding of the IEEE Annual International Conference on Engineering in Medicine and Biology Society, pp. 6107-6110, Minneapolis, United States, Sep. 2009.
[20] S. Chembumroong, S. Cang and A. Atkins, “Elderly activities recognition and classification for applications in assisted living,” Proceeding of Expert Systems with Application, Vol. 40, No. 5, pp. 1662-1674, April 2013.
[21] S. S. Intille, F. Albinali and, S. Mota, “Design of a Wearable Physical Activity Monitoring System using Mobile Phones and Accelerometers,” Proceeding of the IEEE Annual International Conference on Engineering in Medicine and Biology Society, pp. 3636-3639 , Sep. 2011.
[22] L. Bao and S. S. Intille, “Activity recognition from user annotated acceleration data,” Proceeding of Pervasive, Vol. 3001, pp. 1-17, Vienna, Austria, April 2004.
[23] F. A. b. Abdullah, M. F. P. Negara, A. Sayeed, and S. M. Choi, “Classification Algorithms in Human Activity Recognition using Smartphones,” International Journal of Computer and Information Engineering, pp. 77-84, June 2012.
[24] N. Ravi, N. Dandekar, and P. Mysore, “Activity Recognition from Accelerometer Data,” Proceeding of the Twentieth National Conference on Artificial Intelligence, Pittsburgh, Pennsylvania, July 2005.
[25] G. Bieber, A. Luthardt, and C. Peter, “The Hearing Trousers Pocket – Activity Recognition by Alternative Sensors,” Proceeding of the Coference on Pervasive Techologies Related to Assistive Environments, No. 44, Crete, Greece, May 2011.
[26] T. S. Saponas, J. Lester, and J. Fogarty, “iLearn on the iPhone: Real-Time Human Activity Classification on Commodity Mobile Phones,” University of Washington CSE Technical Report, Washington, 2008.
[27] N. Gyorbiro, A. Fabian, and G. Homanyi, “An Activity Recognition System For Mobile Phones,” Springer Magazine of Mobile Networks and Applications, Vol. 14, pp. 82-91, Feb. 2009.
[28] L. Sun, D. Zhang, and B. Li, “Activity Recognition on an Accelerometer Embedded Mobile Phone with Varying Positions and Orientations,” Proceeding of International Conference of Ubiquitous Intelligence and Computing, pp. 548-562, Brisbane, Australia, July 2010.
[29] T. Brezmes, J. L. Gorricho, and J. Cotrina, ”Activity Recognition from Accelerometer Data on a Mobile Phone,” Proceeding of the 10th International Work-Conference on Artificial Neural Networks, pp. 796-799, Heidelberg, Berlin, 2009.