Pattern recognition of an implicitly given target

摘要

Classic correlation filters for detection and location estimation of a reference require explicit information aboutthe object to be recognized. In this work we assume that a reference signal is not available. However, we supposethat the target is placed at unknown coordinates in a noisy reference image. Optimal correlation filters withrespect to signal-to-noise ratio and peak-to-output energy for detection and localization of a target embeddedinto an input scene are derived. Computer simulation results obtained with proposed filters are presented andcompared with those of common correlation filters.