MASINT fusion of multispectral, hyperspectral & kinematic phenomenology

The evolution of Electro-Optical (EO) technology has continuously advanced towards minimizing a sensor´s size, weight and cost while simultaneously increasing key performance metrics such as resolution and range. Frequently, these dichotomous objectives make for a trade space which is difficult to resolve. Fortunately, the use of unresolved multispectral and hyperspectral imagery fused with other Measurement And Signature Intelligence (MASINT) such as target kinematic features presents an opportunity to extend a sensor´s useful classification/discrimination range without requiring additional hardware innovation. In this paper, we will provide an overview of multispectral, hyperspectral and kinematic based MASINT phenomenologies which are available for extraction and exploitation. Additionally, we will review feature and information fusion from the perspective of Maximum Likelihood, Naïve Bayes and Bayesian Belief Networks in the context of MASINT Information Fusion.