Berehulyak O. Information technologies of image processing on the base of two-parameter triangular norm and fuzzy logic

The dissertation is devoted to development of information technologies of digital image transformation and analysis based on triangular norms theory and fuzzy logic. Two-parameter t-norm, is constructed and substantiated in the dissertation. The connection between fuzzy logic and quantitative evaluation of image generalized contrast through triangular s-norms is established. S-type fuzzy membership function of real positive power is proposed. A number of contrast enhancement methods on the base of fuzzy logic are improved. Method of two-parameter fuzzy membership function optimization on the base of fuzzy entropy maximum principle is proposed. The fuzzy logic based methods for noise removal from images which uses multiplicative image model are constructed. Methods for object straight edges and lines recognition on grayscale images which uses Hough transform, fuzzy logic and analyze image local contrast are improved. Developed methods confirmed their efficiency at construction of systems for automate image processing and analysis at non-distractive testing of goods and materials and software for automation of beam moving trajectory creation at electron beam welding process.