A novel technique for the blind deconvolution of an image from an unknown blurring function is presented. This iterative technique consists of two steps. We first estimate the blurring function and then use this result to estimate the original image. The problem is posed in terms of successive interlinked energy minimizations that are mapped onto two Hopfield neural networks.
© 1994 Optical Society of America
Original Manuscript: November 12, 1993
Published: April 15, 1994
R. J. Steriti and M. A. Fiddy, "Blind deconvolution of images by use of neural networks," Opt. Lett. 19, 575-577 (1994)