現今穿戴式裝置蓬勃發展，已經是在運動中量測心率的主流。然而穿戴式裝置本身存在著一些會令使用者感覺不適的問題，例如運動後排汗造成接觸裝置的皮膚產生過敏現象或是身體被束縛等身體不適感。因此，為了解決穿戴式裝置所面臨的問題，近年來有了許多非接觸式的心率量測研究。現今相關的非接觸式心率量測研究大多都是建構在個人電腦上，本論文將非接觸式心率量測系統成功實現在由ARM與GPU所組成的嵌入式系統平台上，並且可以提供日間的運動量測與夜間的健康照護兩種量測方式。本論文結合光流法、色度演算與適應性濾波等方法來提高準確度。日間運動方面，平均絕對誤差為5.48 bpm(beat per minute)，標準差為6.66 bpm。夜間量測方面，平均絕對誤差為1.94 bpm，標準差為2.37 bpm。此外，本系統的畫面更新率為60 FPS(Frame per second)，並且經過實驗發現在60 FPS下平均絕對誤差為5.52 bpm，而30 FPS下平均絕對誤差則高達26.81 bpm，由此得出，60 FPS有著比30 FPS更高的準確度。經由12名受測者的實驗後可以證實本論文所提出的方法具有不低於現今非接觸式心率量測相關研究的準確度。

Nowadays, wearable devices are thriving and are the mainstream of heart rate measurement. However, the wearable device itself has some problems that may cause discomfort to the user. For example, sweaty skin that contact with the device may cause allergies, or physical discomfort such as the body being bound. Therefore, in order to solve the problems above, there have been a lot of researches in non-contact heart rate measurement in recent years. Most of the studies are PC-based. In this paper, we successfully implement the non-contact heart rate measurement system on an embedded system platform consisting of ARM and GPU. This system provides daytime exercise monitoring and nighttime health monitoring. We use optical flow, CHROM, and adaptive filter to improve the accuracy. In daytime exercise the mean absolute error is 5.48 bpm (beats per minute) and standard deviation is 6.66 bpm. In nighttime measurement, the mean absolute error is 1.94 bpm and the standard deviation is 2.37bpm. Besides, the frame rate of our system is 60 FPS(Frames per second). The mean absolute error of 60 FPS is 5.52 bpm, and the mean absolute error of 30 FPS is 26.81 bpm, which concluded that the accuracy is higher when the frame is 60 FPS. Experiments by 12 subjects confirm that the proposed method has higher accuracy than the current studies on non-contact heart rate measurement.

[1] Y.-T. Zhang and C. Y. Poon, "Health informatics: Unobtrusive physiological measurement technologies," IEEE Journal of Biomedical and Health Informatics, vol. 17, pp. 893-893, Sep. 2013.
[2] "快樂動 - 健康體能 - 衛生福利部國民健康署_肥胖防治網," 衛生福利部國民健康署, [Online]. Available: http://obesity.hpa.gov.tw/TC/HealthBodyContent.aspx?id=4&chk=f5a87186-f2be-4dee-a373-da07cab16c05. [Accessed 27 Jan. 2018].
[3] D. Munro, "Credit Suisse Says Wearable Tech 'The Next Big Thing'," Forbes, [Online]. Available: http://b-i.forbesimg.com/danmunro/files/2013/05
/CSDevicesFinal.png. [Accessed 03 Nov. 2016].
[4] A. Sikdar, S. K. Behera, D. P. Dogra and H. Bhaskar, "Contactless vision-based pulse rate detection of Infants Under Neurological Examinations," in Engineering in Medicine and Biology Society (EMBC), 2015 37th Annual International Conference of the IEEE, Milan, Italy, 2015.
[5] L. Feng, L.-M. Po, X. Xu, Y. Li and R. Ma, "Motion-Resistant Remote Imaging Photoplethysmography Based on the Optical Properties of Skin," IEEE Transactions on Circuits and Systems for Video Technology, vol. 25, pp. 879-891, Oct. 2014.
[6] Y.-C. Lin, G.-Y. Lin and Y.-H. Lin, "Development of a FPGA-based contactless pulse rate detection system," in Intelligent Green Building and Smart Grid (IGBSG), 2016 2nd International Conference on, Prague, Czech Republic, 2016.
[7] Y.-C. Lin and Y.-H. Lin, "Step Count and Pulse Rate Detection Based on the Contactless Image Measurement Method," IEEE Transactions on Multimedia, pp. 1-1, Jan. 2018.
[8] T. Tamura, Y. Maeda, M. Sekine and M. Yoshida, "Wearable Photoplethysmographic Sensors—Past and Present," Electronics 2014, pp. 282-302, Apr. 2014.
[9] Z. Zhou, F. Shi and W. Wu, "Learning Spatial and Temporal Extents of Human Actions for Action Detection," IEEE Transactions on Multimedia, vol. 17, pp. 512-525, Feb. 2015.
[10] G. Yu, N. A. Goussies and J. Yuan, "Fast Action Detection via Discriminative Random Forest Voting and Top-K Subvolume Search," IEEE Transactions on Multimedia, vol. 13, pp. 507-517, Mar. 2011.
[11] W. Xu, Z. Miao and X.-P. Zhang, "A Hierarchical Spatio-Temporal Model for Human Activity Recognition," IEEE Transactions on Multimedia, vol. 19, pp. 1494-1509, Feb. 2017.
[12] B. D. Lucas and T. Kanade, "An iterative image registration technique with an application to stereo vision," Proceedings of Image Understanding Workshop, pp. 121-130, Apr. 1981.
[13] G. R. Bradski and A. Kaehler, Learning OpenCV: Computer Vision with the OpenCV Library, O'Reilly Media, 2008.
[14] P. Viola and M. . J. Jones, "Robust Real-Time Face Detection," International Journal of Computer Vision, vol. 57, pp. 137-154, May. 2001.
[15] D. L. Jones, "IIR Filter Structures," [Online]. Available: https://cnx.org/contents/2NeOfN6f@2/IIR-Filter-Structures. [Accessed 27 Jan. 2018].
[16] G. d. Haan and V. Jeanne, "Robust Pulse Rate From Chrominance-Based rPPG," IEEE Transactions on Biomedical Engineering, vol. 60, pp. 2078-2886, Jun. 2013.
[17] R.-Y. Huang and L.-R. Dung, "A motion-robust contactless photoplethysmography using chrominance and adaptive filtering," in Biomedical Circuits and Systems Conference (BioCAS), 2015 IEEE, Atlanta, GA, USA, 2015.
[18] "Polar H7 Website," Polar, [Online]. Available: https://www.polar.com/us-en/products/accessories/h7_heart_rate_sensor. [Accessed 27 Jan. 2018].