Regularised iterative deconvolution algorithms for image restoration based on a Topkis-Voinott approach

This paper addresses a regularized deconvolution technique used on image restoration by an iterative Topkis-Voinott approach. In order to suppress background noise during restoration, a Lagrange parameter is introduced for regularizing the deconvolution produces. A mathematics model is described in this paper and the artificial and real microscope restored images are compared to other algorithms with different noise level. The experiment proves, by giving a reasonable PSF, this algorithm is robust.