Integrated Data Experimentation and Fusion System (IDEFS)

Adaptive data fusion techniques offer widespread potential for application in defense and space arena, ranging from computer vision based target recognition, coastal ocean sensing to geographical information systems. In an intuitive level, data fusion reflects human perception and decision-making capability in taking more than-one source of information into account. A unified approach for designing decision support systems that support data fusion can make an enormous difference in the ease and efficiency of the decision making. The proposed approach for this SBIR effort involves the development of an intelligent adaptive data fusion system using soft computing paradigms including neural networks, fuzzy logic and genetic algorithms. The key components of this approach involves in developing (i) a unified data representation scheme, (ii) a set of aw data into the representation scheme, (iii) methodologies to iteratively develop intelligent data fusion systems based upon soft-computing approaches, (iv) methodologies to extract qualitative and quantitative knowledge out of the soft computing models, (v) methods that support multiple model integration, and (vi) experimentation and design rationale capture mechanisms.