Multiple-object detection with a chirp-encoded joint transform correlator

Abstract

We investigate the performance of a chirp-encoded joint transform correlator in the presence of multiple input objects. We show that, for an input scene containing multiple targets, the chirp-encoding technique focuses the desired cross correlations between the reference signal and the input targets and the undesired self-correlations between the targets in the input scene in separate output planes. The output of the chirp-encoded joint transform correlator is mathematically analyzed for an input scene containing multiple targets. Both the linear joint transform correlator and the nonlinear joint transform correlator in the presence of multiple input targets are considered. For the nonlinear joint transform correlator, the chirp-encoding focuses the higher-order correlation terms, including the higher-order terms of the self-correlations between the targets in the input scene onto separate output planes. The separation requirements of the conventional and the chirp-encoded joint-transform correlator in the presence of multiple input targets are discussed. Computer simulations and experimental results of the chirp-encoded joint transform correlator for a scene containing multiple input targets are presented. The results are compared with a conventional joint transform correlator for an input scene containing multiple targets.