Abstract
Effects of the threshold value on the multiple-target detection performance of a modified amplitude-modulated joint transform correlator are studied on the basis of computer simulations, using human fingerprint and face images. Correlation outputs are quantitatively measured by means of a primary-to-secondary peak ratio and a peak-to-correlation-deviation ratio. The simulation results reveal that the detection of high-contrast targets can be optimized by using a low-threshold. In the case of low-contrast targets, the optimization depends on spatial-frequency contents and noise.

This publication has 9 references indexed in Scilit: