A novel method for computing an optimal stopping time for a nonlinear image restoration technique is discussed in this paper. The stopping criterion is determined using an iterative strategy that minimizes the signal-to-noise ratio (SNR), defined as the ratio between the covariance of the noise, and that of the signal. It is based on the assumption that, if the ideal signal and noise are uncorrelated, then the same must be reflected in their corresponding estimates.
Despite a couple of discrepancies in the text, the paper is very well written, and the discussion is clear. The authors’ argument is mathematically complete, and clearly demonstrates a novel nonlinear technique to determine the stopping criterion in image restoration. The examples clearly illustrate the method, and highlight its performance in comparison to the conventional SNR method.