Improved iterative blind image deconvolution

Abstract

The simple technique of iterative blind deconvolution of two convolved functions has been improved in this chapter. The proposed improvement imposes some new constraints to the iterative algorithm making the result more accurate and more visually appealing. The proposed scheme has also been equipped with a stopping criterion ensuring the convergence of the algorithm which was missing in the original work. As a result, one need not look into the image produced after every iteration to decide the termination of the algorithm. Simulations indicate that the restored images of the proposed version of the scheme are more close to the true image.