Recent advances in automatic radar target identification

Abstract

The author considers a class of algorithms that are applicable to the detection and identification of low-observable targets. These algorithms, which are based on adaptive forms of sequential hypothesis testing techniques are shown to be effective in minimizing the number of measurements required to perform a detection or identification decision. In addition, the author describes the implementation and evaluation of target classification based on structural descriptions derived from polarimetric backscatter measurements. Practical scenarios where the characteristics of the measured signal may be significantly altered by an addition to or deletion from the standard target operating configuration are considered. In these cases, an approach based on target structure is capable of localizing its effects on the classification or interpretation of the target.<>