Wavelet Denoising Methods: VisuShrink, SUREShrink, BayesShrink and Beyond

This program is developed on the WaveLab802 platform for signal denoising applications. It implements multiple advanced denoising methodologies including VisuShrink, SUREShrink, and BayesShrink techniques. The implementation allows users to configure noise variance parameters and select appropriate wavelet bases according to specific requirements, enabling optimized denoising performance. The code architecture supports comprehensive parameter customization, where users can adjust threshold calculation methods and wavelet decomposition levels based on signal characteristics. The VisuShrink method implements universal thresholding using sqrt(2*log(n)) scaling, while SUREShrink employs Stein's Unbiased Risk Estimate for adaptive threshold selection. BayesShrink utilizes Bayesian framework for minimum mean squared error estimation. This versatile denoising toolbox effectively handles various signal types including audio, images, and other data formats, providing high-quality noise reduction through wavelet coefficient thresholding and reconstruction algorithms. The modular design allows for easy integration of additional wavelet families and thresholding strategies, making it suitable for diverse signal processing applications where noise removal and feature preservation are critical.