Title :
Forward-Looking Sonar Target Tracking and Homing From a Small AIV
Author :
Cobb, J. Tory ; Schulz, Bryan ; Dobeck, Gerald
Author_Institution :
Naval Surface Warfare Center, Panama, FL
Abstract :
Neutralization of hazardous obstacles and targets in coastal areas is a labor intensive, dangerous task for human divers. Automated detection, tracking, and classification of bottom and volume targets from autonomous underwater vehicles (AUV) will enable remote neutralization of hazardous objects in much less time than current operations. In order to meet tight power and processing constraints, detection, tracking, and classification algorithms must be fast enough to execute on small onboard computers. This paper describes a real-time algorithm for tracking and homing on volume targets for a small AUV. The algorithm was implemented onboard a Nekton 4 7/8 in. diameter AUV outfitted with a blazed array forward-looking sonar. Performance results of the algorithm in off-line bench tests and from a field demonstration at AUVfest in Keyport, WA are presented
Keywords :
image classification; object detection; object recognition; oceanographic techniques; sonar tracking; target tracking; underwater vehicles; 0.875 inch; Nekton 4 AUV; autonomous underwater vehicles; blazed array forward-looking sonar; coastal areas; forward-looking sonar target tracking; hazardous objects; hazardous obstacles; hazardous targets; target classification; target detection; target homing; volume targets; Classification algorithms; Humans; Object detection; Sea measurements; Sonar; Target tracking; Testing; Underwater tracking; Underwater vehicles; Vehicle detection;
Conference_Titel :
OCEANS, 2005. Proceedings of MTS/IEEE
Conference_Location :
Washington, DC
Print_ISBN :
0-933957-34-3
DOI :
10.1109/OCEANS.2005.1639948