POST-PROCESSING OF JOINT-TRANSFORM CORRELATOR OUTPUTS WITH THE KOLMOGOROV-SMIRNOV STATISTICAL TEST

The Kolmogorov-Smirnov statistical test has been adapted as a template matching algorithm for detecting a binary target in a grey-scale scen e. The algorithm defines a measure of similarity between a template an d a region of the scene. This measure of similarity is then applied to post-processing of joint transform correlator outputs in a pattern re cognition process by computing the product between the measure of simi larity and the intensity out of the correlator at the neighborhood of a possible detection peak. (C) 1998 Elsevier Science B.V.