心血管疾病已成为了人类的第一大杀手, 心率作为表征心血管状况重要指标, 其临床及日常监测对预防心血管疾病引起的猝死具有重要意义。设计了一个基于视频的非接触式运动状态下多人心率实时检测系统, 通过人脸检测与人脸跟踪技术实现多人实时准确检测跟踪, 并采用感兴趣区域自适应提取与ASF滤波算法, 抑制了IPPG信号中的运动伪差, 实现了高信噪比脉搏波信号提取。采用医用指夹式血氧仪进行对比实验, 二者的均方根误差为4.1280BPM、相关系数为0.812。Bland-Altman分析结果显示, 大多数点分布在95%的置信区间内, 表明设计的心率监测系统与标准指夹式测量结果具有较好的一致性。

Cardiovascular disease has become the first killer of human beings. Heart rate is an important indicator of cardiovascular status. Its clinical and routine monitoring is of great significance in preventing sudden death caused by cardiovascular disease. A real-time and non-contact heart rate detection system based on video for multi-people in motion is designed. The accurate detection and tracking in real time is realized by face detection and face tracking technology, adaptive region extraction and ASF filtering algorithm are used to suppress motion artifact in the IPPG signal for achieving high signal-to-noise ratio. Contrast experiments are performed using a medical finger-clip oximeter. The root mean square error of the two is 4.1280 BPM and the correlation coefficient is 0.812. The Bland-Altman analysis shows that most of the points are distributed within the 95% confidence interval, indicating that the heart rate monitoring system designed is in good agreement with the standard finger-clip measurement results.