Image deconvolution using a micro genetic algorithm

Abstract Image deconvolution is addressed using a modified Genetic Algorithm for the error minimization. Numerical examples are presented to demonstrate the algorithm's robustness. Specifically, noise levels and various regions of support are investigated to verify the approach to the blind deconvolution problem of two positively constrained images. Good results were obtained using the algorithm in recovering images from their convolutions without any a priori information, other than the maximum bounds on their regions of support.