Geometric and Temporal Characterization of Battlefield Smoke and Dust by Multispectral Digital Image Analysis

摘要

The objective of this research is to design and demonstrate an automated technique for: the identification of smoke and dust types observed in simulated battlefield environments; the derivation of the geometry and external dynamics of these features; and prediction of the trajectory, expansion, and dispersion rates with time. The premise is that the use of currently operational passive multispectral image collection systems - with output that is functional in a digital processing mode - will reveal information that either cannot be obtained from data acquired by conventional analog techniques such as photography or at least not as quickly and conveniently for the purpose of effectiveness. An extension of this effort will be to correlate this information with particulate and transmissivity measurements in order to accurately quantify the cloud surface into values that are feasible as input to transport and diffusion modeling. (Author)