Learning Horizontal Connections In A Sparse Coding Model Of Natural Images

Image prior models based on sparse and redundant  representations are attracting more and more attention in the field of image restoration. The conventional sparsity-based methods enforce sparsity prior on small image patches independently. Unfortunately, these works neglected the contextual information  between sparse representations of neighboring image patches . Sparse coding of image patches extracted from the whole image can be seen as filtering of the image with a set of filters. The sparsity concept of natural images comes from the early work on transform-domain techniques, such as discrete cosine transform (DCT) and discrete wavelet transform (DWT).When applying a transform to natural images, a few coeffi-cients represent the principal components of image structure. In this paper, we utilize the contextual information of local patches (denoted as context-aware sparsity prior) to augment the   performance of sparsity-based restoration method. In addition , a unified framework based on the markov random fields model is anticipated to tune the local prior into a global one to deal with  arbitrary size images. An iterative numerical solution is presented to solve the joint problem of model parameters estimation  and sparse recovery. Finally, the experimental results on image denoising and super-resolution demonstrate the effectiveness and robustness of the proposed context-aware method