COMPARATIVE PERFORMANCE ANALYSIS OF ADAPTIVE MULTISPECTRAL DETECTORS

The fully adaptive hypothesis testing algorithm developed by Reed and Yu for detecting low-contrast objects of unknown spectral features in a nonstationary background is extended to the case where the relative spectral signatures of objects can be specified in advance. The result ing background-adaptive algorithm is analyzed and shown to achieve rob ust spectral feature discrimination with a constant false-alarm rate ( CFAR) performance. A comparative performance analysis of the two algor ithms establishes some important theoretical properties of adaptive sp ectral detectors and leads to practical guidelines for applying the al gorithms to multispectral sensor data. The adaptive detection of actua l man-made artifacts in a natural background is demonstrated by proces sing multiband infrared imagery collected by the Thermal Infrared Mult ispectral Scanner (TIMS) instrument.