For mobile, landscape view is recommended.
Many aspects of the underwater environment interfere with the detection, characterization, and recovery of military munitions. Except at very shallow sites, munitions underwater are difficult to access. Conditions interfere with the ability of sensors to detect and characterize munitions causing remediation to be more difficult. The SERDP Munitions Response Program Area has been interested in exploring how best to detect and classify submerged munitions using acoustic methods for many years.
This year’s SERDP Project of the Year was headed by Dr. Steven Kargl from the Applied Physics Laboratory at the University of Washington. Dr. Kargl and his project team developed a Gaussian mixture models (GMM) classification scheme and collected low frequency and high frequency synthetic aperture sonar (SAS) data from objects within an acoustically hard environment. GMM is a probabilistic model for representing normally distributed subpopulations.
Dr. Kargl extended previous SERDP efforts that measured acoustic responses from a collection of inert munitions, scientific targets, and clutter items. The central hypothesis was that the environment and the scattering geometry within that environment can alter an object’s acoustic response.
The research validated models for a selection of targets, developed a GMM classification scheme, and collected low-frequency and high-frequency SAS data from objects within an acoustically hard environment. Validated models provided simulated data for training and testing of the GMM classification scheme. Data collected in the acoustically hard environment was used to test the robustness of the GMM classification scheme. A software module for the GMM classification scheme has been provided to an ESTCP project and is being merged into an operations/classification package of a towed platform.
Project Team: