PERFORMANCE ANALYSIS OF ROBUST DETECTORS FOR HYPERSPECTRAL IMAGING

When accounting for heterogeneity and non-Gaussianity of real hyperspectral data, elliptical distributions provide reli-able models for background characterization. Through these assumptions, this paper highlights the fact that robust esti-mation procedures are an interesting alternative to classical methods and can bring some great improvement to the detec-tion process. The goal of this paper is then not only to recall well-known methodologies of target detection but also to pro-pose ways to extend them for taking into account the hetero-geneity and non-Gaussianity of the hyperspectral images. Index Terms — hypespectral imaging, target detection, elliptical distributions, M-estimators 1