Iterative statistical approach to blind image deconvolution

Image deblurring has long been modeled as a deconvolution problem. In the l iterature, the point-spread function (PSF) is often assumed to be known exa ctly. However, in practical situations such as image acquisition in cameras , we may have incomplete knowledge of the PSF. This deblurring problem is r eferred to as blind deconvolution. We employ a statistical point of view of the data and use a modified maximum a posteriori approach to identify the most probable object and blur given the observed image. To facilitate compu tation we use an iterative method, which is an extension of the traditional expectation-maximization method, instead of direct optimization. We derive separate formulas for the updates of the estimates in each iteration to en hance the deconvolution results, which are based on the specific nature of our a priori knowledge available about the object and the blur. (C) 2000 Op tical Society of America [S0740-3232(00)00507-X] OCIS codes: 100.1830, 100. 3020, 100.2000, 000.5490, 110.5200.