Synthesis from underwater data: application to the oceanic discrete tomography

Abstract

In this paper, we describe an efficient method to synthesize, using time-frequency methods, underwater acoustic signals adapted to the Discrete Oceanic Tomography (DOT) by using opportunity acoustic sources naturally present in the ocean. This novel approach is based on two essential tools for the extraction of relevant information in DOT context: an auditory modelling to simulate mammalian auditory perception and a time-frequency Viterbi-based detector to estimate some useful instantaneous frequency laws. The results obtained from real underwater data illustrate the interest and the potential of the proposed synthesis method.