Km. Perry et Sj. Reeves, A PRACTICAL STOPPING RULE FOR ITERATIVE SIGNAL RESTORATION, IEEE transactions on signal processing, 42(7), 1994, pp. 1829-1833
Iterative signal restoration is a common and simple approach for recov
ering degraded signals. Unfortunately, iterative algorithms often conv
erge slowly, and the convergence point is usually not the best restora
tion because of noise amplification. Therefore, a stopping rule should
be imposed to maximize the effectiveness of the iterative algorithm.
We demonstrate the value of randomized generalized cross-validation (R
GCV) as a stopping rule for linear iterative restoration algorithms wi
th simple initial conditions and show that it can be used on relativel
y small data sets with confidence. We also illustrate the performance
of the RGCV criterion for various degrees of blurring and noise.