A Smartphone-Based Living Skin Recognition Method Using Remote-PPG

• Main Authors: Shun-Chieh Yu, 游舜傑
• Other Authors: Yuan-Hsiang Lin
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/48550102283814040289

碩士 === 國立臺灣科技大學 === 電子工程系 === 105 === The poor recognition of out-of-hospital cardiac arrest (OHCA) with checking carotid pulse is less than 50% correct by the public. Thus, to assist public for more effective recognition of cardiac arrest and to facilitate patients receiving early resuscitation, we propose a pulse-correlation-deviation (PCD) method to differentiate the living/non-living skin tissue by analyzing the contactless pulse signal. We firstly constructed the identification method on the smartphone, which is ubiquitous and eligible to accord realistic emergency demands. The image sensor on the smartphone was utilized to capture the remote PPG (rPPG) physiological signal and a cross-correlation operation was used to enhance the characteristics of cardiac condition of a living subject. The results show that the proposed PCD method outperforms the existed method (FDR) by providing faster and higher detection rates (95% and 90% for the fixed-holding and hand-holding smartphone experiment, respectively). In the future, the proposed innovation can facilitate the bystander to recognize OHCA in short time (within 8 seconds), and to execute earlier CPR, earlier PAD, and to provide more information for emergency calls to improve the survival rate of OHCA patients.