A pattern recognition algorithm for the blind discrimination of liquid and solid filled munitions

A new pattern recognition algorithm is described for the nondestructive evaluation of stockpiled munitions. The algorithm performs blind classification on munitions to determine if they are liquid-filled (chemical) or solid-filled (conventional) munitions. The algorithm uses attributes of the acoustic spectrum of the munition under inspection to perform the discrimination. Results are presented on the performance of the pattern recognition algorithm on a large set of data collected from chemical and conventional munitions at the US Tooele Army Depot.

[1]  P. S. Lewis,et al.  Techniques for classifying acoustic resonant spectra , 1994, Proceedings of 1994 28th Asilomar Conference on Signals, Systems and Computers.

[2]  P. S. Lewis,et al.  Munitions classification using an acoustic resonance spectroscopic technique , 1993, Proceedings of 27th Asilomar Conference on Signals, Systems and Computers.

[3]  D.N. Sinha.  Acoustic resonance spectroscopy (ARS) , 1992, IEEE Potentials.

[4]  Keinosuke Fukunaga,et al.  Statistical Pattern Recognition , 1993, Handbook of Pattern Recognition and Computer Vision.