This paper describes developments in the Sonar Active BoundaryLoss Estimation project. The aim of the project is to producehigh-resolution bottom backscatter and bottom-loss databases usingcombatant AN/SQS-53C sonar “through the sensor” datacollection. The approach utilizes modified signal processing techniquesdeveloped specifically for naval sonar systems to automatically detectand identify features of interest in the recorded time series of eachbeam. The current implementation is able to process acoustic returnsfrom a variety of sonar setups. For each ping, all beams are processed.Feature parameters are extracted from the sonar signals, automaticallygeo-referenced, and stored in a database along with supporting ship andsonar status information. Unique features inherent to the database,analysis tools, and visualization capabilities include use of MATLAB, acommercially available software package; on-the-fly creation of griddeddata at arbitrary (user-specified) grid cell resolution; ships status,navigation, and sonar setup data; the ability to extract, analyze, anddisplay parameters by azimuth of observation; and the ability togenerate “images” of the ocean bottom along the ship's trackas seen by the actual sonar system. In this paper we present an overviewof the signal and data processing from the sonar beamformer output tothe production of a gridded geo-referenced database and provide severalexamples of data visualization and data filtering which are possiblewith this approach
展开▼
