Automated EO/IR Detection Techniques For Floating Objects Navy SBIR FY2006.2 Sol No.: Navy SBIR FY2006.2 Topic No.: N06-136 Topic Title: Automated EO/IR Detection Techniques For Floating Objects Proposal No.: N062-136-0092 Firm: Arete Associates P.O. Box 6024 Sherman Oaks, California 91413 Contact: Paul Carter Phone: (818) 501-2880 Web Site: www.arete.com Abstract: DDG 1000's Integrated Undersea Warfare system will employ high frequency sonar for floating MLO detection. This sensor responds to a variety of noise sources at the air-sea interface and typically cannot distinguish between different classes of floating objects. However, by cueing an EO/IR sensor to the region of the surface selected by the HF sonar, floating threats can potentially be automatically detected, classified, and presented for viewing to an operator in a way that reduces workload while improving the accuracy of threat evaluation. This proposal addresses critical new technology for the detection and classification of floating objects including mines and MLOs using the DDG 1000's EO/IR sensor suite. The goal of this SBIR program is to augment the MLO detection algorithm suite with the addition of two new software modules. The first is a false alarm rejection module that would screen false alarm video sequences output by the detection module, thereby relieving the operator of the burden of analyzing clearly non-threatening cases. The second, a classification module, contains logic that will assist the man-in-the-loop in judging how likely a detected MLO is a true threat, providing a confidence level for the leading class assignments along with supporting data. Benefits: The benefits of an automated detection and classification system could extend well beyond the DDG 1000 program. Floating objects, not restricted to MLOs, could present hazards to surface vessels of all types. The technology developed under the proposed effort would be valuable as a means of collision avoidance for use in both naval and commercial ship systems. More broadly, the proposed technology could be adapted for use within the automotive industry: automatic pedestrian detection, driver-assistance systems for intersection left-turn conflicts, lane departure detection, and collision warning for adaptive cruise control are just a few of the applications. Thermal infrared sensors, such as those employed in Raytheon's NightDriverTM automotive night-driving system, are extremely well suited for the detection of collision hazards in dark environments and could be greatly enhanced with algorithms capable of classifying animals, pedestrians, bicycles, and vehicles in a high clutter environment. Return