Perceptually-adaptive image super-resolution using statistical methods

Multi-frame image super-resolution makes use of a set of low-resolution images to reconstruct one or more high-resolution images. This paper presents a novel super-resolution algorithm that uses perceptually important content characteristics such as edges, texture, and brightness to improve visual quality. The super- resolution algorithm introduces perceptually-adaptive constraint relaxation to optimize the image for the human vision system. Experimental results show that the super-resolution algorithm improves visual quality both quantitatively and qualitatively when compared with standard techniques.