Topological Data Analysis and Wide Area Detection of Chemical and Biological Contamination MP 39-10

Metron, Inc. and Stanford University propose to design, develop, test and demonstrate topological data analytic algorithms to analyze hyperspectral imagery. We propose to adapt the topological data analytic techniques, including Stanford’s successful Mapper algorithm, to the hyperspectral imagery domain. Using these algorithms we will identify topological and geometric features and properties indicative of chemical and biological contamination. We will leverage Metron’s expertise in Bayesian filters to produce preliminary detection and classification results using the extracted features and properties.