Image Denoising Based on MORF and Minimization Total Variation

A new tight frame called monoscale orthonormal ridgelet frame (MORF) is presented in this paper. The localization principle and the orthonormal ridgelet constructed by Donoho are applied to construct the MORF. And then, It is deployed for image denoising, where a thresholding process for MORF coefficients of the noisy image is carried out firstly. To remove the artifact such as pseudo-Gibbs and to restore sharp discontinuities, while the other structures are preserved, a minimization total variation approach is applied to restore the denoised results. The experiments show that the method has more improvement both in terms of PSNR and visual effect than traditional thresholding method for image denoising.