Parallel blind deconvolution of astronomical images based on the fractal energy ratio of the image and regularization of the point spread function

Abstract

A parallel blind deconvolution algorithm is presented. The algorithm contains the constraints of the point spread function (PSF) derived from the physical process of the imaging. Additionally, in order to obtain an effective restored image, the fractal energy ratio is used as an evaluation criterion to estimate the quality of the image. This algorithm is fine–grained parallelized to increase the calculation speed. Results of numerical experiments and real experiments indicate that this algorithm is effective.