Image Super-Resolution via Dual-Dictionary Learning And Sparse Representation
This is an incremental improvement for image processing applications, enhancing detail recovery in super-resolution.
The paper tackles image super-resolution by decomposing high-frequency details into main and residual components, using dual-dictionary learning and sparse representation to recover them, resulting in improved PSNR and visual perception compared to state-of-the-art methods.
Learning-based image super-resolution aims to reconstruct high-frequency (HF) details from the prior model trained by a set of high- and low-resolution image patches. In this paper, HF to be estimated is considered as a combination of two components: main high-frequency (MHF) and residual high-frequency (RHF), and we propose a novel image super-resolution method via dual-dictionary learning and sparse representation, which consists of the main dictionary learning and the residual dictionary learning, to recover MHF and RHF respectively. Extensive experimental results on test images validate that by employing the proposed two-layer progressive scheme, more image details can be recovered and much better results can be achieved than the state-of-the-art algorithms in terms of both PSNR and visual perception.