Performance evaluation of hyperspectral detection algorithms for sub-pixel objects

摘要

One of the fundamental challenges for a hyperspectral imaging surveillance system is the detection of sub-pixelobjects in background clutter. The background surrounding the object, which acts as interference, providesthe major obstacle to successful detection. Two additional limiting factors are the spectral variabilities of thebackground and the object to be detected. In this paper, we evaluate the performance of detection algorithmsfor sub-pixel objects using a replacement signal model, where the spectral variability is modeled by multivariatenormal distributions. The detection algorithms considered are the classical matched filter, the matched filter withfalse alarm mitigation, the mixture tuned matched filter and the finite target matched filter. These algorithmsare compared using simulated and actual hyperspectral imaging data.