Typically a regression approach is applied in order to identify the constituents present in a hyperspectral image,
and the task of species identification amounts to choosing the best regression model. Common model selection
approaches (stepwise and criterion based methods) have well known multiple comparisons problems, and they do
not allow the user to control the experimet-wise error rate, or allow the user to include scene-specific knowledge
in the inference process.
A Bayesian model selection technique called Gibbs Variable Selection (GVS) that better handles these issues is
presented and implemented via Markov chain monte carlo (MCMC). GVS can be used to simultaneously conduct
inference on the optical path depth and probability of inclusion in a pixel for a each species in a library. This
method flexibly accommodates an analyst's prior knowledge of the species present in a scene, as well as mixtures
of species of any arbitrary complexity. A series of automated diagnostic measures are developed to monitor
convergence of the Markov chains without operator intervention. This method is compared against traditional
regression approaches for model selection and results from LWIR data from the Airborne Hyperspectral Imager
(AHI) are presented. Finally, the applicability of this identification framework to a variety of scenarios such as
persistent surveillance is discussed.