Nonlinear image restoration using a segmentation-oriented expert system

Abstract

The implementation of a truly autonomous expert system for image restoration is discussed. It involves an attempt to combine two unrelated fields, artificial intelligence and image restoration. Another novel aspect of this research is the application of image segmentation to image restoration. The preliminary experimental results are presented. Specific applications include image restoration from multiple versions, restoration from phase information, and restoration from stochastically degraded images.<>