Blind Image Deconvolution via Particle Swarm Optimization with Entropy Evaluation

This study addresses a blind image deconvolution which uses only blurred image and tiny point spread function (PSF) information to restore the original image. In order to mitigate the problem trapping into a local solution in conventional algorithms, the evolutionary learning is reasonably to apply to this task. In this paper, particle swarm optimization (PSO) is therefore utilized to seek the unknown PSF. The objective function is designed according to entropy theorem whose evaluation can distinguish characteristics between a blurred image and a clear image. Finally, the feasibility and validity of proposed algorithm are demonstrated by several simulations; further, its performance is compared with that of another state of the art evolutionary algorithm.