ECG signals are usually corrupted by baseline wander, power-lineinterference, muscle noise, etc. and numerous methods have been proposed toremove these noises. However, in case of wireless recording of the ECG signalit gets corrupted by the additive white Gaussian noise (AWGN). For the correctdiagnosis, removal of AWGN from ECG signals becomes necessary as it affects theall the diagnostic features. The natural signals exhibit correlation amongtheir samples and this property has been exploited in various signalrestoration tasks. Motivated by that, in this work we propose a nonlocalwavelet transform domain ECG signal denoising method which exploits thecorrelations among both local and nonlocal samples of the signal. In theproposed method, the similar blocks of the samples are grouped in a matrix andthen denoising is achieved by the shrinkage of its two-dimensional discretewavelet transform coefficients. The experiments performed on a number of ECGsignals show significant quantitative and qualitative improvement in denoisingperformance over the existing ECG signal denoising methods.
展开▼
