基于BiEWT算法和谱估计的非接触式心率测量研究

Research on non⁃contact heart rate measurement based on BiEWT algorithm and spectrum estimation

传统非接触式心率测量采用快速傅里叶变换和峰值检测方法,存在易受环境噪声影响、测量结果波动较大、精度较低的问题。因此,提出一种二重经验小波变换(Binary Empirical Wavelet Trans⁃form,BiEWT)和功率谱估计(Power Spectrum Estimation,PSE)相结合的非接触毫米波雷达心率测量方法。BiEWT算法融合了自适应分解方法和引入先验知识的经验小波分解方法的优势。通过R-EWT提取去除与噪声相关的多尺度分析(Multiresolution Analysis,MRA)分量,并引入心率频率先验知识进行频点筛选。通过F-EWT进一步抑制残留噪声,重构出心跳信号。利用功率谱估计方法,计算被测人的心率。实验结果表明,该方法比单次高模数的EWT算法,注意力更集中于心率频率区间。比FFT+PD算法MAPE平均下降了1.74%,比WNR+PSE算法MAPE平均下降了1.18%。

Traditional non⁃contact heart rate measurement using fast Fourier transform and peak detection method is easy to be affected by noise,resulting in high fluctuation of heart rate measurement results and low accuracy.Therefore,a non⁃contact millimeter⁃wave radar heart rate measurement method combining Binary Empirical Wavelet Transform(BiEWT)and Power Spectrum Estimation(PSE)is proposed.BiEWT algorithm mainly combines the advantages of adaptive decomposition method and empirical wavelet decomposition method which introduces prior knowledge.R-EWT is used to extract and remove the signal of Multiresolution Analysis(MRA)components related to noise,and make frequency point screening processing by introducing the prior knowledge of heart⁃rate frequency.F-EWT is used to further suppress residual noise and reconstruct pure heartbeat signal.The power spectrum estimation method is used to calculate the human heart rate.Experimental results show that compared with the single high⁃modulus EWT algorithm,this method focuses more on the interval of heart rate frequency.Compared with FFT+PD algorithm,MAPE decreased by 1.74%on average,and compared with WNR+PSE algorithm,MAPE decreased by 1.18%on average.