Sparse Image Reconstruction for Partially known Blur Functions

In this paper, we consider the problem of image reconstruction from the noisy blurred version of an original image when the blurring operator is partially known and the original image is sparse. Using optimization transfer, we derive a novel iterative algorithm in closedform that incorporates both sparseness and partial knowledge of the image. We demonstrate the performance of the algorithm using simulations.

• Type: Conference
• Year: 2006
• Where: ICIP
• Authors: Raviv Raich, Alfred O. Hero
• Keywords: ICIP 2006, Image Processing, Noisy Blurred Version, Original Image