Image Super Resolution Enhancement using IBP Method
Sruthi Mohan · 2014
This proposed paper provide Adaptive Iterative pixel (IBP) method for image super-resolution (SR). This paper presents a novel self-learning approach for SR. This proposed framework can provide Iterative Pixel, which offers excellent generalization in modelling the relationship between images and their associated SR versions. Unlike most prior SR methods, this does not require the collection of training low and high-resolution image data in advance, and do not assume the reoccurrence of image patches within an image or across image scales. With theoretical supports of Deconvolution filter verify that SR framework learns and selects the optimal Iterative Pixel model when producing an SR image, which results in the minimum SR reconstruction error. Evaluate AIBP method on a variety of images, and obtain very promising SR results. In most cases, this method quantitatively and qualitatively outperforms.