Intelligent Multisensor Integration and Fusion

This SBIR project proposes the investigation and development of an advanced intelligent information fusion system using neural networks and expert systems. Our approaches feature the following innovations: 1) A Digent neural network for data fusion problem in the presence of noise is designed to be robust to prescribed noise margins; and 2) A real time knowledge processing system is developed for management and implementation of dynamic control, multiple assignment, and tracking algorithms. The knowledge base that exists in the machine and the knowledge engineering used to capture the expert's skills is explained. The developed expert system can teach and aid less-experienced operators in decision making. A prototyping system will be developed to demonstrate the advanced multisensor integration and fusion techniques to be conducted in this project. The prototyping system will be designed to be user-friendly and easy to use through employing state-of-the art window and object-oriented programming techniques. The developed techniques will be transferred to our commercial product: Intelligent Multisensor Integration and Fusion Systems. The effectiveness of the proposed intelligent multisensor fusion approaches will be demonstrated on a fully integrated prototype module at the end of Phase I. BENEFITS: The developed techniques will result in the commercial product: Intelligent Multisensor Integration and Fusion Systems. Airline pilots, power station operators, air defense tactical operations center commanders, military commanders, manufacturing plant managers and others are the candidates for this project.