Method of target detection in images by moment analysis of correlation peaks

Automatic target detection and recognition in images often is attempted by use of a linear correlation filter (matched filter), whose output is interp reted by a single pointwise detector (detection based on only one point). I examine a technique for significantly improving the performance of this ta rget detection approach by supplementing the pointwise detector with severa l neighborhood correlation peak detectors (detection based on a domain of m any points extending over much of the peak). The neighborhood detectors ext ract peak shape information through a moment analysis of correlation plane peaks. I describe the design of statistically quasi-optimal correlation pea k discriminators based on second-order geometric moments. (C) 1999 Optical Society of America. OCIS codes: 100.5010, 100.4550, 100.1160, 100.2000.