We consider the optimal likelihood algorithm for the estimation of a target
location when the images are corrupted by substitutive noise. We show the
relationship between the optimal algorithm and the sliced orthogonal nonlin
ear generalized (SONG) correlation. The SONG correlation is based on the ap
plication of a linear correlation to corresponding binary slices of both th
e input scene and the reference object with appropriate weight factors. For
a particular case, we show that the optimal strategy is a function of only
the number of pixels for which the gray values in the noisy image match th
e ones of the reference image when the substitutive noise is uniformly dist
ributed. This is exactly what a particular definition of the SONG correlati
on does. (C) 2001 Optical Society of America.