In this paper, a new subband-based classification scheme is developed for c
lassifying underwater mines and mine-like targets From the acoustic backsca
ttered signals. The system consists of a feature extractor using wavelet pa
ckets in conjunction with linear predictive coding (LPC), a feature selecti
on scheme, and a backpropagation neural-network classifier, The data set us
ed for this study consists of the backscattered signals from six different
objects: two mine-like targets and four nontargets for several aspect angle
s. Simulation results on ten different noisy realizations and for signal-to
-noise ratio (SNR) of 12 dB are presented. The receiver operating character
istic (ROC) curve of the classifier generated based on these results demons
trated excellent classification performance of the system, The generalizati
on ability of the trained network was demonstrated by computing the error a
nd classification rate statistics on a large data set, A multiaspect fusion
scheme was also adopted in order to further improve the classification per
formance.