SBIR-STTR Award

An integrated approach for performing fusion using multiple sensors, national assets, and doctrinal information is proposed. The target application of the proposed approach is the strategic relocatable target problem. A solution to this problem requires capabilities beyond current capabilities. A novel multi-stage model-based approach is proposed. Individual technology elements that are in need of further development are identified. A special emphasis is placed on fusion techniques for passive ir and active 3d laser data.

A key technical problem of increasing national importance is the detection, location, and classification of relocatable military targets. This problem arises in a number of important applications including monitoring and targeting land mobile strategic relocatable targets (SRT) and tactical warfare. The advent of new sensor systems including Synthetic Aperture Radar (SAR) and 3D laser radar promise increased information yet presents new problems in information processing and sensor fusion that are beyond the current state of the art. Model-based recognition techniques show great promise in overcoming current limitations. However, model based technology is still relatively immature, will require significant technology development, and its ultimate performance, requirements, and limitations are still unknown. The proposed effort addresses these concerns. The effort includes: a technology development of model baed fusion algorithms for SAR and 3D imagery data; an assessment of the algorithm performance; and an evaluation of the limitations and requirements of model based processing as a function of mission and system parameters. Optional tasks will continue this work but concentrate on CC&D effects. Anticipated Benefits/Potential Commercial Applications - Model based vision for sensor fusion is a new technology that is expected to provide significant improvements in automatic target recognition (ATR), reconnaissance, robotics, medicine, and manufacturing.