A fuzzy shadow feature scheme for radar signal detection

Shadow feature algorithms are used to enhance target detectability in radar systems when the target obscures some portion of a nearby clutter field. This paper proposes and describes a new shadow feature algorithm, the fuzzy shadow length estimation (FSLE) algorithm, to improve the accuracy of shadow length estimation, thereby enhancing the performance of some underlying detection algorithm. The scheme uses fuzzy membership functions to measure the degree to which signal returns beyond the target resemble a shadow and then estimates the shadow length. It can be applied to any detection algorithm using a binary decision rule. In this paper, the FSLE algorithm is applied to maximum likelihood constant false alarm rate (ML-CFAR) detection of a Rayleigh fluctuating target in Weibull clutter. The FSLE algorithm significantly improves target detectability in situations in which a shadow is cast, particularly in high clutter environments