The USC Andrew and Erna Viterbi School of Engineering USC Signal and Image Processing Institute USC Ming Hsieh Department of Electrical Engineering University of Southern California

Technical Report USC-SIPI-360

“Multi-Category Classification of Ground Vehicles Based on the Acoustic Data Using Fuzzy Logic Rule-Based Classifiers”

by Hongwei Wu and Jerry M. Mendel

November 2003

This report summarizes our research that was conducted from July 2002 to July 2003 about the multi-category classification of ground vehicles---heavy-tracked, light-tracked, heavy-wheeled and light-wheeled vehicles---based on the acoustic data that were collected in the normal environmental conditions.

We have proposed three fuzzy logic rule-based classifier (FL-RBC) architectures, one non-hierarchical and two hierarchical, have conducted leave-one-out, leave-two-out and 10-fold cross validation experiments to evaluate the performances of these architectures, and have also compared them to a Bayesian classifier. Our experimental results showed that for the multi-category classification problem, (1) FL-RBCs performed substantially better than the Bayesian classifier, (2) type-2 FL-RBCs performed better than their competing type-1 FL-RBCs, (3) the type-2 non-hierarchical and hierarchical in series architectures performed the best, and (4) the adaptive operational mode gave much better performance than the non-adaptive operational mode.

To download the report in PDF format click here: USC-SIPI-360.pdf (21.7Mb)