论文部分内容阅读
由于心电图(ECG)信号的特点以及在采集过程中所受到的干扰影响,ECG信号去噪已成为ECG信号智能分析的基础。本文在基于小波变换方法的基础上,对阈值参数进行改进,提出了与噪声更加匹配的阈值表达式。利用改进的阈值对离散分解后的小波系数进行处理,通过小波逆变换重构信号,能够更加准确地去除噪声的小波系数,保留更多原信号小波系数。采用MIT-BIH中的数据进行实验,结果表明,改进方法较之现有小波阈值去噪方法,能够达到更好的去噪效果。
ECG signal de-noising has been the basis for the intelligent analysis of ECG signals due to the characteristics of ECG signals and the interference they are subjected to during acquisition. Based on the wavelet transform method, this paper improves the threshold parameters and proposes a threshold expression that matches the noise better. By using the improved threshold to process discrete wavelet coefficients and reconstructing signals through inverse wavelet transform, the noise wavelet coefficients can be removed more accurately, and more original wavelet coefficients remain. Experiments using MIT-BIH data show that the improved method is better than the existing wavelet threshold denoising method to achieve better denoising effect.