A Structured Sub-pixel Target Detector for Hyperspectral Imagery

Structured detectors like the adaptive matched subspace detector (AMSD) have been proposed for sub-pixel target detection in hyperspectral images for years. They model hyperspectral data with a subspace model, which proved to be effective in separating the sub-pixel targets from the background. This paper proposes a structured detector in which selective endmembers are used according to different pixels to ensure that the true composition of endmembers in each pixel is utilized in both the subspace model and the statistical test. Experiments show its better performance for sub-pixel targets detection than current structured detector.