疫情衍生出「零接觸」的醫療服務需求，藉由這些遠距、非接觸式裝置以量測生物醫學訊號是未來醫療科技的發展主向之一；老年人跌倒是公共衛生及其臨床重要的議題之一，偵測跌倒的市場需求始終熱絡。非接觸感測器，不但有利於長期連續監測及個人化醫療，擷取生理訊號、影像判讀並提供臨床可用之參考指標。
遠程光體積描記圖法（Remote Photoplethysmography, rPPG），以非接觸式方式，相較傳統設備上具有高舒適性、高便利性、能配合長時間監控的優點，可多方應用於不同場合。本研究基於POS(Plane Orthogonal to Skin)並以Sub-band優化演算法，並以測試於rPPG的公開資料庫，同時使用其他現有的rPPG方法作為比較，並使用Bland-Altman分析，結果顯示，差值平均d ̅介於[-1.43~0.38]，95%一致性的上下限(d ) ̅± 1.96 *Sd =[ -8.63 ~ 13.53, -9.00 ~ -16.40]，優於其他比較方法。
現行跌到偵測系統大多採配戴感測器方式進行分析，此類方法易將日常活動當作誤判，誤判率較高，也需基於使用者習慣下使用。而許多AI影像辨識技術，大多集中於可見光攝影機的影像領域，對於使用熱影像，解析度極低的熱影像上的技術仍舊不多，熱影像只是呈現人的大致形體，可保留更多個人隱私訊息，且不受環境光源設置影響。
本研究以使用網路視訊攝影機基礎，藉由影像資訊，發展出基於皮膚反射模型開發一盲源分離演算法，該演算法可有效降低光源變化與動作運動干擾影響，穩定提升訊號除平均心率外，轉換成脈動波型(pulsatile waveform)以利視覺化效果及計算即時心率及呼吸分析；本研究核心以非接觸式環境感測器為主，將此延伸應用至兩相關主題，包含: (1) 基於影像感測器開發人臉影像追蹤rPPG盲源分離演算法與(2)基於深度學習於極低解析度熱像辨識人體姿態辨識初步結果。

The measurement of biomedical signals by non-contact devices is one of the main directions for the future development of "zero-contact" medical technology, and falls in the elderly are critical to public health and clinical practice. Non-contact sensors are not only beneficial for long-term continuous monitoring and personalized medicine, but can also capture physiological signals, perform image interpretation, and provide clinically usable reference indicators.
Compared with traditional equipment, Remote PPG has the advantages of high comfort and convenience. This study is based on POS (Plane Orthogonal to Skin) and adopts Sub-band optimization for robust rPPG, This algorithm can effectively reduce the impact of light source changes and motion interference and .tests it on the public database of rPPG while using other existing rPPG methods as a comparison, and using Bland-Altman analysis, the results show that the difference is that the averaged ̅ is between [-1.43~0.38], the upper and lower limits are 95% consistent (d ) ̅±1.96 *Sd = [-8.63~13.53, - 9.00 ~ - 16.40], which is more than the other the comparison method is better.
fall detection systems use contact sensors for analysis can easily lead to misjudgment of daily activities. Many AI technologies are mostly focused on the imaging field of visible light cameras. For the use of thermal imaging, there are not many thermal imaging techniques with extremely low resolution. Thermal images, allowing more personal privacy information, are not affected by ambient light settings, so we use deep learning-based ultra-low resolution thermal images for human pose recognition and show preliminary results

[1] Khanokh B., Slovik Y., Landau D., Nitzan M., “Sympathetically induced spontaneous fluctuations of the photoplethysmographic signal”, Medical & Biological Engineering & Computing, 2004, vol. 42, pp.80-85.
[2] Wenjin Wang∗, Albertus C. den Brinker, Sander Stuijk, and Gerard de Haan, Algorithmic Principles of Remote PPG, IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 64, NO. 7, 2017
[3] Xu, Zhi-Feng & Jia, Rui-Sheng & Liu, Yan-Bo & Zhao, Chao-Yue & Sun, Hong-Mei. (2020). Fast Method of Detecting Tomatoes in a Complex Scene for Picking Robots. IEEE Access. PP. 1-1. 10.1109/ACCESS.2020.2981823.
[4] J. Yang, W. Lu and A. Waibel, Skin-color modeling and adaptation, ACCV98, 1998.
[5] S. L. Phung, A. Bouzerdoum and D.Chai,“A novel skin color model in YCbCr color space and its application tohumanfacedetection,”2002 International Conference on Image Processing., vol 1, pp.289- 292, Sept. 2002.
[6] Qiang Zhu, Kwang-Ting Cheng and Ching-TungWu,”Adaptive learning of an accurate skin-colormodel,”Sixth IEEE International Conference on Automatic Face and Gesture Recognition 2004 Proceedings, pp.37-42, May 2004.
[7] T. S.Caetano and D. A. C. Barone, “A probabilistic model for the human skin color,“11th International Conference on Image Analysis and Processing (ICIAP'01), pp. 279-283, Sept. 2001
[8] S. Bobbia, R. Macwan, Y. Benezeth, A. Mansouri, J. Dubois, "Unsupervised skin tissue segmentation for remote photoplethysmography", Pattern Recognition Letters, 2017.
[9] C.S. Pilz, S. Zaunseder, J. Krajewski, V. Blazek, Local Group Invariance for Heart Rate Estimation from Face Videos in the Wild, The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pp.1254-1262, Salt Lake City, 2018.
[10] Hoffman, W.F.C.; Lakens, D. Public Benchmark Dataset for Testing rPPG Algorithm Performance; 4TU.Centre for Research Data: The Hague, The Netherlands, 2019.
[11]M.-Z. Poh et al., “Advancements in noncontact, multiparameter physiological measurements using a webcam,” Biomedical Engineering, IEEE Transactions on, vol. 58, no. 1, pp. 7–11, Jan. 2011.
[12]G. de Haan and V. Jeanne, “Robust pulse rate from chrominance-based rPPG,” Biomedical Engineering, IEEE Transactions on, vol. 60, no. 10, pp. 2878–2886, Oct. 2013.
[13]Chen, Liang-Chieh, et al. "Encoder-decoder with atrous separable convolution for semantic image segmentation." Proceedings of the European conference on computer vision (ECCV). 2018.
[14] Chollet, François. "Xception: Deep learning with depthwise separable convolutions." Proceedings of the IEEE conference on computer vision and pattern recognition. 2017.
[15] M. Zimmermann, H. Bunke. “Hidden Markov Model Length Optimization For Handwriting Recognition Systems”, Proc. of the 8th International Workshop on Frontiers in Handwriting Recognition, pp. 369–374, 2002.
[16] J. M. Bland and D. G. Altman, “Agreement between methods of measurement with multiple observations per individual,” J. Biopharmaceutical Statist., vol. 17, no. 4, pp. 571–582, 2007.
[17] M. Lewandowska et al., “Measuring pulse rate with a webcam - a noncontact method for evaluating cardiac activity,” in Computer Science and Information Systems (FedCSIS), Federated Conference on, Sept. 2011, pp. 405–410.
[18] M.-Z. Poh et al., “Advancements in noncontact, multiparameter physiological measurements using a webcam,” Biomedical Engineering, IEEE Transactions on, vol. 58, no. 1, pp. 7–11, Jan. 2011.
[19] Niu X S, Han H, Shan S G . 2020. Remote photoplethysmography-based physiological measurement: a survey. Journal of Image and Graphics, 25(11): 2321-2336.
[20] G. de Haan and A. van Leest, “Improved motion robustness of remote-ppg by using the blood volume pulse signature,” Physiological Measurement, vol. 35, no. 9, pp. 1913–1922, Oct. 2014.
[21] W. Wang et al., “A novel algorithm for remote photoplethysmography: Spatial subspace rotation,” IEEE Trans. Biomed. Eng., vol. 63, no. 9, pp. 1974– 1984, Sept. 2016.
[22]B. Jacob, J. Chen, Y. Huang, and I. Cohen, “Pearson correlation coefficient,” in Noise Reduction in Speech Processing. Berlin, Germany: Springer-Verlag, 2009, pp. 1–4.