Virtual experts for imagery registration and conflation

The unique human expertise in imagery analysis should be preserved and shared with other imagery analysts to improve image analysis and decisionmaking. Such knowledge can serve as a corporate memory and be a base for an imagery virtual expert. The core problem in reaching this goal is constructing a methodology and tools that can assist in building the knowledge base of imagery analysis. This chapter provides a framework for an imagery virtual expert system that supports imagery registration and conflation tasks. The approach involves tree strategies: (1) recording expertise on-the-fly and (2) extracting information from the expert in an optimized way using the theory of monotone Boolean functions and (3) use of iconized ontologies to built a conflation method.

Key words: Imagery virtual expert, ontology, knowledge base, rule generation optimization, monotone Boolean function, registration, conflation.