Spectral Ultrasonic Underwater Buried Object Detection and Localization

Acoustic object detection in underwater applications, such as the detection of unexploded ordnances (UXO), is challenging, especially at shallow burial depths or in the presence of other reflections. Therefore, acoustic detection is performed with very short temporal and spectrally very broadband pulses. The received signal strength is measured and displayed over time (time of flight). However, this method cannot detect closely spaced reflections because the pulses overlap in the time domain. In this work, we present spectral signal processing that allows the detection and localization of known object geometries in the presence of interfering reflections. We extend our underwater ultrasound propagation model in this work to include a sediment layer. We apply spectral fingerprinting to detect a known object geometry and localize it in a second step. We evaluate the performance with measurements in a laboratory experiment with an object with varying burial depths and on sediment.