交通事故已為10大死因的其中之一，而異常駕駛行為相當容易引起交通事故，為了識別異常行為，需要建構駕駛行為模型，過去有許多關於建構駕駛行為模型的研究，但是許多需要額外的儀器，實用上造成額外的經濟負擔或不便。隨著穿戴式裝置的普及與目前新款汽車逐漸可以對應可攜裝置界面，同時市面上的穿戴式裝置搭載的感測器如加速度計、陀螺儀與磁力計，令使用穿戴式裝置建構駕駛者行為模型有新的可能性。在本研究中，針對實驗室過去基於模擬環境駕駛資料提出的方法，為了得知該系統方法可否移植到真實車輛駕駛中，系統性的分析其方法在真實駕駛環境與模擬駕駛環境的差異是否會對識別效能造成顯著影響，並證實持續性小震動對智慧手錶之駕駛者行為模型建構方法會造成顯著負面影響，需要用中值濾波器濾除；不會因突發性大震動對識別效能造成顯著影響，同時我們提出了一套藉由車內環境感測器濾除所有震動的演算法，但經實驗證實此方法無效，經研究分析發現在讓兩個感測器讀數投影到同一座標系這件事會導致辨識率下將，才導致濾除演算法無效。

Abnormal driving behaviors can easily cause traffic accidents. To identify abnormal driving behaviors of a people, driving behavior modeling is crucial. Smartwatch becomes more and more common and we can use it to analyze driving behavior. In this research, we analysis whether the difference between the real road driving and the driving simulator will have a significant effect on the method our laboratory proposed that modeling the distribution of the hand-movement feature of the driver obtained from the smartwatch by Gaussian mixture models (GMMs).We prove that our method would have significant negative impact to continuously vibration and we need to use median filter to do data pre-processing. Also, we prove that method won’t have significant impact to suddenly huge vibration. Although in this work, we proposed a filter to filtering all vibration by another car environment sensor reading, but it isn’t work when experiment. We found that align driver behavior sensor and car environment sensor let the accuracy become worse.

[1] WORLD HEALTH ORGANIZATION, “The top 10 causes of death,” World Health Organization, 2011. [Online]. Available: http://www.who.int/mediacentre/factsheets/fs310/en/index1.html.
[2] C.Saiprasert andW.Pattara-Atikom, “Smartphone Enabled Dangerous Driving Report System,” in 2013 46th Hawaii International Conference on System Sciences, 2013, pp. 1231–1237.
[3] K.Igarashi, C.Miyajima, K.Itou, K.Takeda, F.Itakura, andH.Abut, “Biometric identification using driving behavioral signals,” in 2004 IEEE International Conference on Multimedia and Expo (ICME) (IEEE Cat. No.04TH8763), 2004, vol. 1, pp. 65–68.
[4] C.Miyajima et al., “Driver Modeling Based on Driving Behavior and Its Evaluation in Driver Identification,” Proc. IEEE, vol. 95, no. 2, pp. 427–437, Feb.2007.
[5] A.Wahab, Chai Quek, Chin Keong Tan, andK.Takeda, “Driving Profile Modeling and Recognition Based on Soft Computing Approach,” IEEE Trans. Neural Networks, vol. 20, no. 4, pp. 563–582, Apr.2009.
[6] H.Qian, Y.Ou, X.Wu, X.Meng, andY.Xu, “Support Vector Machine for Behavior-Based Driver Identification System,” J. Robot., vol. 2010, pp. 1–11, 2010.
[7] A.Riener andA.Ferscha, “Supporting Implicit Human-to-Vehicle Interaction : Driver Identification from Sitting Postures,” in Proceedings of the First Annual International Symposium on Vehicular Computing Systems. July 22-24, 2008, Dublin, 2008, pp. 1–10.
[8] R.Chen, M. F.She, X.Sun, L.Kong, andY.Wu, “Driver recognition based on dynamic handgrip pattern on steeling wheel,” Proc. - 2011 12th ACIS Int. Conf. Softw. Eng. Artif. Intell. Netw. Parallel Distrib. Comput. SNPD 2011, pp. 107–112, 2011.
[9] “In-vehicle Wearable Integration to Accelerate Convergence; Global Penetration in New Cars to Exceed 90% by 2019,” ABI Research, 2014. .
[10] Bhawna Kohli, “Smartwatch Market by Type Product (Extension, Classic & Standalone smartwatch) and Application (PA, Wellness, Medical/ Health, Sports) - Global Opportunity Analysis and Industry Forecast, 2013 - 2020,” 2015.
[11] C.Yang, D.Liang, andC.Chang, “A novel driver identification method using wearables,” in 2016 13th IEEE Annual Consumer Communications & Networking Conference (CCNC), 2016, pp. 1–5.
[12] H.Eren, S.Makinist, E.Akin, andA.Yilmaz, “Estimating driving behavior by a smartphone,” in 2012 IEEE Intelligent Vehicles Symposium, 2012, no. 254, pp. 234–239.
[13] Z.Chen, J.Yu, Y.Zhu, Y.Chen, andM.Li, “D3: Abnormal driving behaviors detection and identification using smartphone sensors,” in 2015 12th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON), 2015, pp. 524–532.
[14] B.-L.Lee, B.-G.Lee, andW.-Y.Chung, “Standalone Wearable Driver Drowsiness Detection System in a Smartwatch,” IEEE Sens. J., vol. 16, no. 13, pp. 5444–5451, Jul.2016.
[15] F.Lin, D.Liang, andE.Chen, “Financial ratio selection for business crisis prediction,” Expert Syst. Appl., vol. 38, no. 12, pp. 15094–15102, 2011.
[16] K.He, J.Sun, andX.Tang, “Guided Image Filtering BT - link.springer.com,” Link.Springer.Com, vol. 6311, no. Chapter 1, pp. 1–14, 2010.
[17] L.Wang, H.Ning, T.Tan, andW.Hu, “Fusion of Static and Dynamic Body Biometrics for Gait Recognition,” IEEE Trans. Circuits Syst. Video Technol., vol. 14, no. 2, pp. 149–158, Feb.2004.
[18] D. A.Reynolds, T. F.Quatieri, andR. B.Dunn, “Speaker Verification Using Adapted Gaussian Mixture Models,” Digit. Signal Process., vol. 10, pp. 19–41, 2000.
[19] A. J. L. H. and R. V.S. O. H. Madgwick, “Estimation of imu and marg orientation using a gradient descent algorithm,” in 2011 IEEE International Conference on Rehabilitation Robotics (ICORR), 2011.