Quality assessment of acoustic color signatures

Abstract

Recent years have witnessed increasing research by the minehunting community into the use of wideband, widebeam, low frequency sonar, typically below 50 kHz, whose purpose is to excite and observe an object's structural response. These signatures exhibit uniqueness allowing mines to be more clearly distinguished from clutter objects. Considerable effort has been directed toward studying the scattering physics, but less work has been done to optimize the signal processing used for extracting signatures from data collected at sea under realistic conditions. Specifically, the tradeoff between signals, sources of noise, and signal processing parameters is not well understood. For example, a long observation interval can capture structural acoustic effects occurring later in time than the geometric scattering from an object, at the expense of increasing the interference from sea floor reverberation. Relationships such as this are described, and metrics are suggested for choosing the most appropriate data colle...