The Study Of The Methods Of Blind Image Restoration Based On Sparse Regularization

The study of the image restoration not only has important theoretical significance,but also has urgent demand in life. In theory, the image restoration is that we takeactions to remove or reduce the image’s degradation during the imaging, and its goal isto deal with the degraded images and make the images tend to the ideal images. But inpractical application, the imaging system’s point spread function (PSF) is unknown ingeneral. Only can we use the degraded images, as well as the part of the priorinformation about the imaging system. Then we directly estimate the point spreadfunction and the real images, which is called the blind image restoration. In life theblind image restoration has a wide range of applications.Based on many scholars’ research, and in order to solve the ill-posed problems inthe blind image restoration，we respectively use a regularization function and theLaplace operator to solve the problem which is that the point spread function is difficultto be determined accurately during blind image restoration. At the same time, theexperiments are conducted and we obtain good results.The main contents of this dissertation are expressed as follows:1. This paper describes the image restoration’s theory and other related theoreticalknowledge in detail. At the same time it describes the ill-posed problem during theimage restoration, the sparse representation, and the use of regularization methodsto solve the ill-posed problems in detail.2. In order to solve the ill-posed problems during the blind image restoration, we use akind of scale invariance and sparse regularization function. In the third chapter ofthis paper we describe and analyze its algorithm detailedly, and conduct someexperiments.3. We use the Laplace operator to solve the ill-posed problem further, and at the sametime eliminate the estimation of the point spread function (PSF) during the blindimage restoration. So the problems are simplified greatly. In the fourth chapter ofthis paper we describe its theory and algorithms in detail, and at the same time weconduct some experiments.4. In the fifth chapter, we summarize the use of several blind image restorationmethods, and prospect future research.