Design and Evaluation of Hyperspectral Algorithms for Chemical Warfare Agent Detection

摘要

Remote sensing of chemical warfare agents (CWA) with stand-off hyperspectral imaging sensors has a wide range of civilian and military applications. These sensors exploit the spectral changes in the ambient photon flux produced by either sunlight or the thermal emission of the earth after passage through a region containing the CWA cloud. The purpose of this paper is threefold. First, to discuss a simple phenomenological model for the radiance measured by the sensor in the case of optically thin clouds. This model provides the mathematical framework for the development of optimum algorithms and their analytical evaluation. Second, we identify the fundamental aspects of the data exploitation problem and we develop detection algorithms that can be used by different sensors as long as they can provide the required measurements. Finally, we discuss performance metrics for detection, identification, and quantification and we investigate their dependance on CWA spectral signatures, sensor noise, and background spectral variability.