A comparative study of signal restoration techniques for the reduction of random noise
D. Stevens, Mrinal Kumar Das · 2002
This paper presents a comparative study of three different techniques for restoration of signals corrupted with Gaussian noise. These include classical frequency domain Wiener filtering, a wavelet based technique, and a novel discrete cosine transform (DCT) data compression based technique. We show that wavelet analysis provides the optimal improvement in the signal to noise ratio with the DCT technique performing nearly as well.