基于人工智能的非接触式心率检测系统  

Artificial Intelligence-Based Contactless Heart Checks System

作　　者：李成杰 葛书含 马笑笑 LI Chengjie;GE Shuhan;MA Xiaoxiao(Jiangsu Normal University Kewen College,Xuzhou Jiangsu 221132,China)

机构地区：[1]江苏师范大学科文学院,江苏徐州221132

出　　处：《信息与电脑》2023年第22期164-166,共3页Information & Computer

基　　金：2023年江苏省大学生创新创业训练项目“非接触式健康检测智能终端”(项目编号:202313988001Y)。

摘　　要：针对现如今人们工作压力大、患心血管及心肺相关疾病的潜在风险更高等问题,提出基于人工智能的非接触式心率检测方法,使大众更加便捷掌握自身心率状况。基于改进粒子滤波结合AdaBoost与CamShift的混合算法,定位与检测人脸,基于皮肤光学模型构造了色度通道商,利用相邻两帧像素通道的商消除直流分量,计算出能反映人体生理信息的1维信号,经过平滑滤波处理及傅里叶分析获得心率,从而反映人们的心率实时状况。In response to the current issues of high work-related stress and increased risks of cardiovascular and respiratory diseases,this article introduces a non-contact heart rate monitoring method based on AI.This method allows the general public to conveniently track their own heart rate status.It utilizes a hybrid algorithm combining improved particle filtering with AdaBoost and CamShift to locate and detect faces.Leveraging a skin optical model,it constructs a chromaticity channel ratio by using the ratio of pixel channels between adjacent frames to eliminate the DC component.This process calculates a 1D signal that reflects physiological information,which is then subjected to smoothing filtering and Fourier analysis to obtain the heart rate,thereby providing real-time information about people’s heart rate condition.

关 键 词：人工智能 非接触式 人脸检测 皮肤光学模型 傅里叶分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]