At the previous Approximation Theory XII meeting, I discussed some preliminary work with the Applied Physics Institute at Western Kentucky University in using multiwavelets to provide an objective analysis of gamma-ray spectrum generated from fast neutron bombardment of objects, for the purpose of identifying the elemental composition of the object. The method discussed at the time worked moderately well with the limited amount of data provided, but subsequent use with data sets of different compounds and with different detectors brought to light serious flaws with its implementation.
This talk will illustrate those issues and will address how they have been corrected in the current implementation. The algorithm has potential commercial application in analyzing gamma-ray, X-ray, fluorescence, etc. spectra, and is protected by provisional patent. It is currently being used in the grant-funded development of a prototype device for analyzing the active/inactive status of small (less than 2 kg in weight) unexploded ordnance.
Available at: http://works.bepress.com/bruce_kessler/65/