Salient region detection and sparse representation based super-resolution approach for chromosome images

Abstract

Nowadays, there has been an increasing interest in using sparse representations for image processing especially super-resolution. This research presents a new approach to sparsity-based image super-resolution in chromosome image, based on sparse signal representation. Research on image statistics suggests that image patches can be well-represented as a sparse linear combination of elements from an appropriately chosen dictionary. Used by this observation, we seek a sparse representation for each patch of the low-resolution input, and then use to generate the high-resolution output. Using Yang's super-resolution method, Bicubic Interpolation method's results are better than super-resolution method's results in chromosome images. Our proposed method's, successful results were obtained by using Toboggan Filter is providing to extract image regions and learning dictionary from divided patches appropriate size.