An ECG Signal Denoising based on VMD and Undecimated Wavelet Trannsform

The ECG signal shows the any abnormalities in the heart conditions. The diagnosis of the ECG signal is very much difficult for a trained physician as these signals are non-stationary. The signal acquisition and its processing can be affected by the noise and will lead to a negative interpretation of the results and the analysis. The common technique that is used for the reduction of noise is the DWT thresholding because of its multiresolution analysis which is much more efficient than the Fourier based filtering. Here the proposed ECG signal denoising is based on VMD. The DWT thresholding is applied to each of the VMFs that are generated by the VMD. In this work the residue is not considered and the signal reconstruction is done by weighted summation where the weights are determined the CLS algorithm. Another denoising method based on UWT in which SWT is used for filtering the noisy signal is also proposed in this work.