Image restoration of multiple noisy images by use of a priori knowledge of the anisoplanatic point-spread function

We propose and simulate an iterative maximum-likelihood image reconstruction method that uses multiple images taken through the turbulent atmosphere under photon-limited conditions. The method assumes that the point-spread function corresponding to each image has been measured by a guide star or other technique. Although a guide star is still required, wave-front correction is achieved through postprocessing instead of by the mechanical methods currently used with adaptive optics. A restored image of greatly increased quality was found to be produced after only 10 iterations. Additional iterations continued to reduce the mean-square error.