Restoration of blurred images by convolution of arbitrary point spread function

Abstract

The holographic inverse filter produced in 19831 has proved successful in the deblurring of linear blurred and defocused images. The purpose of this paper is to further demonstrate its new use, i.e., the restoration of blurred images resulting from convolution of arbitrary point spread function. The characteristic features of the holographic inverse filter are (1) it is generated by using a weak reference beam instead of a sandwich filter; (2) to compensate the narrow dynamic range of the recording material, a mask is inserted in front of the holographic plate to attenuate the object and reference beams at the same rate; (3) a noise suppression technique is used with the holographic inverse filter to eliminate noise and achieve high image clarity.