Blind Deconvolution of Still Images using Recursive Inverse Filtering

This thesis presents a novel blind deconvolution technique for the restoration of linearly degraded images without explicit knowledge of either the original image or point spread function. The technique applies to situations in which the scene consists of a nite support object against a uniformly grey background. This occurs in applications such as astronomy, and medical imaging. The only information required are the nonnegativity of the true image and the support size of the original object. A novel supportnding algorithm is proposed for situations in which the exact object support is unknown. The restoration procedure involves equalization of the blurred image using a convex cost function. The performance of the technique for truncated equalizer parameters, and in the presence of noise are examined analytically. The new approach is experimentally shown to be more reliable and to have faster convergence than the existing nonparametric nite support blind deconvolution methods.