Optimal stopping condition for iterative image deconvolution by a new orthogonality criterion

The stopping condition is a common problem for non-regularized deconvolution methods. We introduce an automatic procedure for estimating the ideal stopping point based on a new measure of independence, checking an orthogonality criterion of the estimated signal and its gradient at a given iteration. We provide an effective lower bound estimate than the conventional ad-hoc methods, proving its superiority to the others at a wide range of different noise models.