Image Blind Restoration And Spatial Resolution Improvement Research

In reality, images are degraded in some degree because of various reasons, so It is an important research topic to improve image quality and deblur image to obtain sharp one. In this thesis, the research is based on image blind restoration and improving image spatial resolution, which aims at researching image blind restoration methods to estimate original undistorted image by means of known or unknown partial prior knowledge about blurred image itself and (or) imaging system, also, researching super resolution reconstruction, we try to break through the restriction of hardware devices and improve the image spatial resolution in some degree so as to obtain better detailed information of image.This dissertation includes the following parts: image degradation models; out-of-focus blurred image restoration; total variation minimizing frequency iterativeness blind image restoration; image spatial resolution improvement. Research products are as follows:(1) In view of out-of-focus blur, we propose two different methods to estimate PSF. According to out-of-focus pill box degradation model, the method of PSF estimation based on parameter-curve manner is proposed by using of Laplacian gradient, square gradient and error square measure; In terms of the deficiency of traditional out-of-focus model, the estimation processing of image PSF is designed according to circle symmetry of out-of-focus blur and through inverse Abel transformation. Finally, increment Wiener filtering is used to effectively restore blurred image by these two methods.(2) Based on the total variation minimizing blind restoration method, in virtue of...