Automatic target recognition and geo-location for side scan sonar imagery

Ocean and lake floor survey is most efficiently conducted by active sonar owing to its favorable propagation range under water relative to other sensing modalities. Range however must be traded for resolution, yielding sonar return imagery that challenges both operator and machine to match returns to specific objects. In the case of certain objects, time-range extent in echo imagery may be exploited by machine learning detection and classification methods. This talk describes the success of such a method when applied to data collected from a low cost autonomous platform. Further capabilities such as decision support and detection geolocation are discussed. [This material is based upon work supported under Air Force Contract No. FA8721-05-C-0002 and/or FA8702-15-D-0001. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the U.S. Air Force. Presentation may contain distribution limited material.]