Techniques for classifying acoustic resonant spectra

A second-generation nondestructive evaluation (NDE) system that discriminates between different types of chemical munitions is under development. The NDE system extracts features from the acoustic spectra of known munitions, builds templates from these features, and performs classification by comparing features extracted from an unknown munition to a template library. Improvements over first-generation feature extraction, template construction and classification algorithms are reported. Results are presented on the performance of the system on a large data set collected from surrogate-filled munitions.<<ETX>>