<title>Chemical cloud mapping and identification using convex cone analysis of AIRIS multispectral imaging data</title>

Abstract

The US Army is currently developing a Chemical Imaging Sensor for the wide-area detection and identification of chemical warfare agents. As part of this effort, LWIR multispectral imaging data was collected with the Adaptive Infrared Imaging Spectroradiometer. Laboratory experiments were conducted using chemical agent simulants, including DMMP and DIMP at different vapor concentration levels. The resulting 14-band images were analyzed by a least squares approach as well as by Convex Cone Analysis (CCA). Both analyses demonstrate the potential of multispectral chemical imaging for wide-area standoff detection. In addition, it is shown that CCA is capable of detecting the presence of spectrally active chemical vapors and estimating their absorbance spectrum without any prior knowledge about the presence or composition of the chemical vapors.