Heart rate monitoring is important for diagnosis and prevention of physiological diseases. The method of measuring heart rate by infrared sequence image has high accuracy. But shaking is a factor to affect the measurement result. Even if a slight shaking will create a big measurement error. In this paper, we propose an automatic capture algorithm of heart rate measurement through the real-time tracking and capturing of the face monitoring area, so the disturbance caused by human body shaking is overcome. Compared with the traditional method<sup>1</sup> , the measurement accuracy of the tests with shaking is improved about 6% higher based on our algorithm.
Blood oxygen saturation is an important physiological parameter to measure human health. The oxygen saturation can be measured in real time through an optical lens with high accuracy. However, when there is the light reflection of the measured area, the accuracy of the measurement results will fluctuate, leading to the instability of the measurement results. In this paper, we improved the original algorithm by detecting and removing the highlight region. The experimental results show that the accuracy of the improved method is higher than that of the traditional method in the case of specular reflection and the stability of the measurement results is enhanced.