Authors: Michael B. Brady*, National Geospatial-Intelligence Agency, Lynn Albin, National Geospatial-Intelligence Agency, Paul Barnes, Naval Research Lab, Tim Cox, Naval Research Lab
Topics: Coastal and Marine, Geographic Information Science and Systems, Remote Sensing
Keywords: geospatial intelligence, GEOINT, maritime domain awareness, automatic identification system, synthetic aperture radar
Session Type: Poster
Start / End Time: 8:00 AM / 9:40 AM
Room: Lincoln 2, Marriott, Exhibition Level
Presentation File: No File Uploaded
Spaceborne observations used to detect ocean going ships support maritime domain awareness by providing a means to understand and monitor maritime activities that could impact security, safety, the economy, or environment. Methods continue to evolve to leverage the growing number of spaceborne assets and observations to understand the maritime environment and provide decision support. Methods that integrate multiple sources of spaceborne observations are of particular interest to enhance detection capabilities. In the current effort, we collect the European Space Agency’s publicly available Sentinel Ground Range Detected Synthetic Aperture Radar (SAR) imagery from the Alaska Satellite Facility, and process it using the Ocean Watch Laboratory (OWL) algorithm. The OWL algorithm automatically detects ships from SAR imagery. Then we compare the SAR-based ship detections with ship detections from spaceborne automatic identification system (AIS) observations using a correlator algorithm. The output is Keyhole Markup Language (KMZ) files of detected ships, which can be readily viewed using GoogleEarth. From the results we are able to: 1) corroborate SAR-based ship detection using the AIS; 2) detect possible ships not detected from SAR imagery to identify potential OWL improvements; 3) detect false positives or “dark targets” from SAR-based detections with no corresponding AIS observations. Correlating SAR imagery with AIS data advances MDA by improving ship detection capabilities in support of maritime activity understanding and decision support. Approved for public release, 19-298.