Research On The Video Images Super-resolution Enhancement Technology

As one of the hotspots of image processing research, Video Image Enhancement adopts super-resolution reconstruction to enhance image. Super-resolution reconstruction means refining super-resolution image from low-resolution image to meet practiacal demands. It has advantages of lower cost, effective image processing of pre-detected material, resolution enhancement, elimination of fuzziness and noise brought by optical system. Meanwhile, super-resolution has a wide application in the fields of public security, remote sensing, preservation of cultural relics, medical imaging.This paper illustrartes the theory of image super-resolution, give an overview of basic of reconstruction, modeling, algorithm, and has a study on image registration from the aspects of principle, steps and methodology. As two main methods, Optical flow and block match algorithm are introduced. An improved method of over step size of first step out of overall three steps, which makes registration error avoidable and has a good evaluation of small scope dynamics, has been proposed. As a result, image registration becomes more accurate and efficient.Super-resolution reconstruction algorithm mainly refers to an improved algorithm of posterior based on the optimal solution of variable parameter, which makes super-resolution image evaluation an optimal choice. Adaptive process of Object function resolves the problem of insufficient image evaluation resulted from dierect setting of traditional algorithm, which improves the image clarity, detail protection and overall smoothness.The Simulation are based on improved search strategy and reconstruction algorithm, a quality evaluation of post-super-resolution image shows desirable results, In the end, a prospect of image super-resolution as well as issue need to be addressed is proposed.