摘要
光电容积脉搏波采集过程中存在基线漂移和高频噪声会给后续人体生理参数的测量带来困难,因此消除噪声干扰是准确进行相关生理参数测量的关键问题。提出一种结合经验模态分解和奇异值分解的去噪方法。该方法采用经验模态分解将光电容积脉搏波信号分解为若干个固有模态函数,通过功率谱密度判断代表基线漂移信息的固有模态函数获得基线漂移曲线;使用奇异值分解处理光电容积脉搏波信号中的高频噪声,针对传统的差分谱法无法准确识别奇异值有效阶次的不足,提出加权能量贡献率的方法选取奇异值的有效阶次。实验结果表明,该方法能有效消除光电容积脉搏波信号中的基线漂移和高频噪声,这对光电容积脉搏波信号检测精度的提高具有重要意义。
Baseline drift and high frequency noise during the photoplethysmography(PPG)collection process make the follow-up measurement of human physiological parameters difficult. As a result,denoising becomes a key for accurately measuring related physiological parameters. A new method that combines the empirical mode decomposition(EMD)and the singular value decomposition(SVD)is proposed in this paper. In this method,the PPG signal is decomposed into several intrinsic mode functions(IMFs)by using EMD,and the baseline drift curve can be obtained by using the power spectral density to determine the IMFs representing baseline drift information. SVD is used to process the high frequency noise in PPG signals. In allusion to the disadvantage that the conventional differential spectral method cannot accurately recognize effective order ranks of singular values,the Percent of Contribution to Total Energy(PCTE)method is put forward to select effective order ranks of singular values.The experimental results show that the method can effectively remove baseline drift and high frequency noise from PPG signals,which has great significance for the improvement of PPG signal detection precision.
出处
《现代电子技术》
北大核心
2018年第4期65-69,74,共6页
Modern Electronics Technique
基金
国家自然科学基金面上项目(61170171)
江苏省"六大人才高峰"项目(2010-WLW-006)
江苏省高校科研成果产业化推进项目(JHB2011-45)
江苏高校优势学科建设工程资助项目~~
