Fuzzy logic classification in image processing

Image comparison and classification systems based on conventional two-valued logic are mostly not flexible enough while evaluating a ''human-decision''-based classification result. Using such systems, it is additionally very hard to represent the experts knowledge in order to calculate the result in video real time because of the amount of data to be processed: In this paper image pre-processing methods are investigated in order to evaluate global image describing parameters which are presented to the classifier. Several classification methods are examined and compared with respect to a fast and human-decision-based evaluation of the degree of similarity between images and patterns.