抄録
A technique is described for restoring digital images using an innovative spatially variant inverse filter that corrects for the image blurring caused when the point spread function (PSF) of a real imaging system with a lens changes its size or shape. The PSFs are estimated for each small local area using a recursive method, and an inverse filter is constructed for each area. The cost function for determining the PSFs is a function of the thickness of the phase-only-synthesis image-edge line. To obtain high-quality images, a new image sharpening technique is applied to the restored images. To reduce noise in the images, the perturbation method is introduced in our work. Simulation showed that this technique restores digital images better than conventional techniques.
