Using Contactless Facial Image Recognition Technology to Detect Blood Oxygen Saturation

Abstract

Since the outbreak of COVID-19, as of January 2023, there have been over 670 million cases and more than 6.8 million deaths worldwide. Infections can cause inflammation in the lungs and decrease blood oxygen levels, which can lead to breathing difficulties and endanger life. As the situation continues to escalate, non-contact machines are used to assist patients at home to monitor their blood oxygen levels without encountering others. This paper uses a general network camera to capture the forehead area of a person’s face, using the RPPG (remote photoplethysmography) principle. Then, image signal processing of red and blue light waves is carried out. By utilizing the principle of light reflection, the standard deviation and mean are calculated, and the blood oxygen saturation is computed. Finally, the effect of illuminance on the experimental values is discussed. The experimental results of this paper were compared with a blood oxygen meter certified by the Ministry of Health and Welfare in Taiwan, and the experimental results had only a maximum error of 2%, which is better than the 3% to 5% error rates in other studies The measurement time was only 30 s, which is better than the one minute reported using similar equipment in other studies. Therefore, this paper not only saves equipment expenses but also provides convenience and safety for those who need to monitor their blood oxygen levels at home. Future applications can combine the SpO2 detection software with camera-equipped devices such as smartphones and laptops. The public can detect SpO2 on their own mobile devices, providing a convenient and effective tool for personal health management.

Contactless Blood Oxygen Saturation Estimation from Facial Videos Using Deep Learning

“…A and B are experimentally evaluated coefficients that are determined by identifying the line of best fit between the ratios of the red and blue channels and the SpO2 estimated by a ground truth device. Following Equation ( 2), remote SpO2 measurement with an RGB camera was further validated in [21][22][23]48,50]. However, only two methods used deep learning and were tested on a public benchmark dataset [48,49].…”

“…Over the last decade, several contactless SpO 2 measurement approaches have been proposed. Researchers have used a variety of cameras, from infrared cameras [ 14 ] and high-quality monochrome cameras equipped with special filters [ 15 , 16 , 17 , 18 ] to off-the-shelf webcams [ 19 , 20 , 21 , 22 , 23 ], to estimate SpO 2 by capturing subtle light intensity changes on the face. Deep learning techniques have achieved state-of-the-art performance for the remote measurement of physiological signs such as HR [ 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 ] and RR [ 36 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 ].…”

Exploring the Feasibility of Vision-Based Non-Contact Oxygen Saturation Estimation: Considering Critical Color Components and Individual Differences

“…Especially in infectious disease situations such as COVID-19, it can increase the risk of infectious diseases, as you have to use a device within someone else's use. As these problems and the risk of infection have become more prominent, non-contact oxygen saturation measurement methods have been actively studied in response [6,7].…”