Wavelet Shrinkage Models for Denoising of Biomedical Signals

In this thesis an efficient denoising models for biological signals using shrinkage function and shrikage based adaptive filters are addressed. Signal denoising, an important problem of signal processing aims to find an estimate of the original signal from the noisy signal. There exist various transform-domain algorithm based on performing orthogonal transformation, modifying transforms coefficients and inverse transforming the modified transform coefficients. In this thesis, transform- domain signal denoising algorithm are equipped with wavelet subband dependent threshold and prioritized shrinkage. Hence they are made more adaptive to the local variations of the signals, thus improving the overall denoising efficiency. The main objective of this work is to develop and implement a denoising model for reducing the noises present in biological signals such as ECE, EEG, PCG, Pulse, EMG etc., and to eliminate the power line frequency interference in these signals. The major work of this thesis is developement of three new shrinkage functions namely hyper, modified hyper and subband adaptive shrinkage.