Abstract—Infrasound signals of interest (SOI) have been collected from various sources including boats, aircraft and personal watercraft during a recent field exercise. The designed experiment specifically located infrasound sensors to gather the SOIs and process them using a developmental methodology. This custom pre-processing, and exploitation suite allows for the extraction of SOIs from experimental data sets that might be sub-optimal due to the presence of noise or interfering (undesired) sources. The on-going development of the infrasound sensor suite and the SOI signal processing and analysis is supported by this work.
Index Terms—Feature vector, infrasound, infrasound analysis, infrasound array, infrasound classification, principal component analysis, PCA.
William. W. Arrasmith is with the Department of Engineering Systems, Florida Institute of Technology, Melbourne, FL. 32901 USA (e-mail: warrasmi@fit.edu)
Everett. R. Coots is with the Harris Corporation, Government Communication Systems Division, 1025 W. NASA Blvd. Melbourne, FL. 32904 USA (e-mail: ecoots@harris.com).
Eric. A. Skowbo is with the Northrop Grumman Corporation, Research and Technology and Military Aircraft Systems, 2000 W. NASA Blvd. Melbourne, 32904 USA (e-mail: eric.skowbo@ngc.com).
John. V. Olson is with the Geophysical Institute, University of Alaska Fairbanks, (e-mail: jvo@gi.alaska.edu).
Barry Webster is with the Department of Engineering Systems, Florida Institute of Technology, Melbourne, FL. 32901 USA (e-mail: bwebster@fit.edu).
[PDF]
Cite: Everett R. Coots, William W. Arrasmith, Eric A. Skowbo, John V. Olson, and Barry Webster, "Target Detection and Classification From Sub-optimal Experimental Data Using Principal Component Analysis and Feature Vector Masking," International Journal of Modeling and Optimization vol. 6, no. 2, pp. 110-118, 2016.