A Blind Image Restoration Method Based on PSF Estimation

Abstract

In order to improve the quality of the restored image, a blind image restoration method is proposed, by estimating the blur function of the imaging model. Firstly, the parameters of the Gaussian point spread function (PSF) of the observed image are estimated. Through Wiener filter image restoration algorithm, multiple error-parameter curves are generated at different parameters. According to these curves, the size and standard deviation of PSF may be estimated. Then, utilizing the estimated PSF, the blurred image is restored through Wiener filter. Experimental results show that this PSF estimation method can estimate the parameters of Gaussian PSF accurately, and justify the fact that PSF estimation plays an important part in image restoration. The PSNR of the restored image has the highest PSNR around the real PSF, and the PSNR decreases when the estimated PSF is far away from its real value.