一种利用手指图像测量人体心率的方法

Measurement of human heart rate using finger pictures

在智能手机上,仅利用手机摄像头采集手指颜色的变化图像,实现人体心率的测量.采用图像频谱分割法对采集的手指图像进行滤波,结合3种像素包含的信息及包含的信息的差异性,利用3种像素功率密度谱加权计算出光电容积脉搏波.通过相隔点检测波峰法检测光电容积脉搏波信号,测出人体心率.将测量结果与标准仪器测量结果进行对比,平均误差为1.82%,与真实值的相关系数大于90%,证明了从手指图像测量出人体心率参数的可靠性.将该测量方法应用到手机上,能极大地方便慢性病人对自己健康状态的监控.

On the smart phone,a phone camera is used to capture finger images to achieve heart rate measurements.First,we apply the spectrum division method to filter the collected finger images,and then use the weighted power density spectrum of the three pixel to calculate the PPG（PhotoPlethysmoGraphy）, based on the information and its differences contained in the three pixels.Then we apply the peak detection method based on the detection of the separated PPG signal to measure the heart rate.Comparing the measured results and the standard results,the average error is 1.82%,and the correlation of the true value is greater than 90%,indicating the reliability of heart rate measurement on finger images.The application of the method can greatly facilitate the convenience of chronic disease monitoring.