System Comparison Procedures for Automatic Target Recognition Systems

摘要

Estimating the performance of an automatic target recognition (ATR) system interms of probability of successful target identification involves extensive image collection and processing, which can be very time-consuming and expensive. We investigate the Wald sequential test for the difference in two proportions as a sample size-reducing alternative to ranking and selection and the classical method of comparing binomial confidence intervals. The test is modified for the multiple pairwise comparison of four systems, and is applied to actual data to compare different configurations of the Moving and Stationary Target Acquisition and Recognition (MSTAR) System.