ECG signal denoising method and system based on classification matching and fractional diffusion

The invention provides a ECG signal denoising method based on classification matching and fractional diffusion, comprising the following steps: S1, inputting an ECG signal to be processed; 2, carryingout sliding classification match on the ECG signal, and extracting a similar data matrix of each reference signal segment; S3, denoising each similar data matrix by shrinking transform coefficients;4, returning the signal segments in each denoised similar data matrix to the original position, and obtaining a preliminary estimation signal by weighted summation; S5. further denoising the obtainedpreliminary estimated signal by fractional diffusion to obtain the final denoised ECG signal. At the same time of removing the high-frequency noise of the electrocardiogram signal, the invention can further remove the low-frequency noise of the same frequency band in the electrocardiogram signal and protect the detail characteristic information in the electrocardiogram signal. 本发明提供种基于分类匹配和分数阶扩散的心电信号去噪方法，包括以下步骤：S1.输入待处理的心电信