A methodology for neural network based classification of marine sediments using a subbottom profiler

A seafloor classification methodology, based on a parameterization of the reflected signal in conjunction with neural network classifiers, is evaluated through computer simulations. Different subbottoms are represented by a stratified model. Using a computer simulation program, these subbottoms were insonified by a chirp signal (2.5-4.5 kHz). Physical parameters are extracted from the simulated acoustic signals. A two stage feature selection method and a radial basis function network classifier are presented. The results indicate that this approach is a promising way for practical, realizable solutions to the problem of remote seafloor classification with a subbottom profiler.