“…The proposed algorithm exhibits superiority over traditional algorithms and recently proposed ones in terms of visual quality, Peak Signal to Noise Ratio (PSNR) and Mean Absolute Error (MAE). Yang et al [6] have introduced the concept of examplesaided redundant dictionary learning into the single-image super-resolution reconstruction, and have proposed multiple dictionaries learning scheme inspired by multitask learning, in order to avoid a large training patches database and obtain more accurate recovery of HR images. Chen et al [7] have designed a new sparsitybased algorithm for the classification of hyper spectral imagery.…”
“…The proposed algorithm exhibits superiority over traditional algorithms and recently proposed ones in terms of visual quality, Peak Signal to Noise Ratio (PSNR) and Mean Absolute Error (MAE). Yang et al [6] have introduced the concept of examplesaided redundant dictionary learning into the single-image super-resolution reconstruction, and have proposed multiple dictionaries learning scheme inspired by multitask learning, in order to avoid a large training patches database and obtain more accurate recovery of HR images. Chen et al [7] have designed a new sparsitybased algorithm for the classification of hyper spectral imagery.…”
“…To distinguish the image sub-blocks with different characteristics, clustering is commonly applied and similar sub-blocks are grouped in each cluster. Learning a local dictionary from a cluster of similar image samples has been proven to be efficient in image denoising [14] and image superresolution reconstruction [47] . However, it has not been widely applied in CS.…”
Section: Local Dictionary Trained From Clustered Image Patchesmentioning
confidence: 99%
“…In addition, the learned dictionary from a particular group of image samples has been proved to be effective in many applications of image processing such as denoising [14] and superresolution [47] . For BCS, there exist significant differences among various image sub-blocks.…”
Section: Local Dictionary Trained From Clustered Image Patchesmentioning
“…Clustering techniques are then added into the sparse representation based SISR scheme in order to employ the priors of the extracted raw patches. E.g., Yang et al [15] learn numerous dictionaries from several groups of patches produced by K-means algorithm and validate its superiority experimentally.…”
Recent years has witnessed an increasing interest in handling the issue of single image super-resolution (SISR) reconstruction. Many researches have demonstrated that the sparse representation based approaches, which rely on the ideal that image patches are assumed to have brief representations when expressed in the proper learned dictionaries, can lead to the stateof-the-art performance. The SISR quality via these algorithms depends strongly on whether the utilized dictionaries can describe the potential high resolution image well, therefore dictionaries learning (DL) is of the greatest significance among the procedures. In this paper we determine to utilize a kind of geometric structure based image patches clustering method combined with K-SVD (a DL algorithm) to obtain a better set of geometric dictionaries. To make further progress, we introduce one regularization term, consistency of gradients, into the framework. This term can better preserve the edges in the reconstructed image, making them sharper. Extensive experiments on natural images indicate that our proposed method outperforms some state-of-the-art counterparts in terms of both numerical indicators and visual quality.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.