Rotation/Scale Invariant Hybrid Digital-Optical Correlator System for Automatic Target Recognition

A distortion invariant hybrid digital-optical correlator system is developed for automatic target recognition. A MACH filter is synthesized using a set of training images. Using this MACH filter, rotation invariance from −30° to 30° and scale invariance from 75% to 140% has been achieved. Effect of distortion on targets embedded in additive and multiplicative noise has been studied.