| Due to the extensive applications of super-resolution image reconstruction and the rapid development of electronic technology, super-resolution image reconstruction has now become a hot topic.This technology is implemented by certain technique for low-quality,low-resolution images to obtain a high-resolution image.Blind image super-resolution is particularly important in practice such as completely unknown or partially known the information of image system. In this paper, for two most common types of image degradation cases including motion degradation and defocus, the problem of estimating the degradation parameters are first studied, and then reconstruction method for single image and image sequence are implemented.Firstly, a comprehensive description of the super-resolution reconstruction, existing super resolution technique and the application areas of this technique are discussed. Secondly, two methods to estimate the degradation parameters about the motion and defocus degradation are presented. One method is called automatic error-parameter curves, and the other is auto-correlation curves on projection. A lot of experiments show that these two methods can quickly and accurately estimate the parameters for motion and defocusing degradation.In the case of getting an estimation of the degradation parameter, the spreading wave equation method is adopted for motion degradation to super resolution reconstruction, and the Lucy-Richardson (LR) iterative is used to get a higher resolution image for the defocus degradation. The experimental results show that the proposed methods provide satisfactory results of the reconstruction.In view of a number of images or image sequences, firstly estimate parameters for each degradation image with above-mentioned methods. Secondly get the displacement relative to the reference frame. Thirdly Projection Onto Convex Sets(POCS) is used for super resolution reconstruction. A summary and the future researches are pointed out in the last part. |
PDF Full Text Request