ABSTRACT
An extensive bibliography of image restoration filters and techniques can be found in a paper by S. A. Rajala and R. J. P. De-Figueiredo. Most of these techniques share a common disadvantage, namely a priori knowledge of the original image is needed. Once the degrading function is found the restoring process can commence. The restoration is performed in two steps: Phase correction and amplitude boost. The amplitude boost cannot be determined in a unique way. In the unrealistic case of no noise the gain function is obvious, the inverse of H. As noise is always present, using the inverse filter will strongly enhance the noise and an unintelligible image will result. The delineated procedure can be applied automatically to any defocused image. No operator intervention is needed. The algorithm allows the filter to be calculated and adapted to the particular image to be reconstructed.
