Image Deblurring Regularized by Wavelet Probability Shrink
Zhiming Wang · 2011
An image deblurring algorithm based on wavelet probabiLity shrink regularization is proposed. Denoise and deblur were alternatively executed by least square approximate and probabiLity shrinkage. After several iterations of deblur, probabiLity shrink based on stationary wavelet transform (SWT) were used once for denoising. Experimental results show that proposed algorithm obtained better results on several benchmark images than classical regularization techniques such as wavelet soft shrink, TV, or even non-local TV proposed more recently.