Tetrolet regularization and learning for single frame image super-resolution

Abstract

A single frame image super-resolution reconstruction technique is proposed with two stages contains tetrolet regularization and tetrolet learning. In the first stage, the tetrolet regularization is used to estimate an initial highresolution image. In the second stage, the tetrolet coefficients at finer scales of the estimated high-resolution image are learned locally from a set of high-resolution training images. Finally the fusion of tetrolet reconstruction produces the super-resolution image. Experimental results demonstrated that the proposed method outperforms state-of-the-art super-resolution methods in terms of PSNR index and visual quality. © Springer-Verlag 2013.