Image denoising is an important image processing task, both theory and application. One of the most useful methods of image denoising is to using wavelet shrinkage. In the method, threshold selection is of great importance. A new threshold function based on the empirical wavelet coefficients has been proposed in this paper. The result of experiments proves our new wavelet shrinkage threshold function method gives a better PSNR (Peak Signal-to-Noise Ratio) to Donoho’s and Wang’s.
展开▼