Wykorzystanie metod sztucznej inteligencji do wizyjnej analizy właściwości warstw wierzchnich

[b]Abstract[/b]. The article presents a selected area of research on the possibility of automatic prediction of material properties based on the analysis of digital images. Original, holistic model of forecasting properties of surface layers based on a multi-step process that includes the selected methods of processing and analysis of images, inference with the use of a priori knowledge bases and multi-valued fuzzy logic, and simulation with the use of finite element methods is presented. Surface layers characteristics and core technologies of their production processes such as mechanical, thermal, thermo-mechanical, thermo-chemical, electrochemical, physical are discussed. Developed methods used in the model for the classification of images of the surface layers are shown. The objectives of the use of selected methods of processing and analysis of digital images, including techniques for improving the quality of images, segmentation, morphological transformation, pattern recognition and simulation of physical phenomena in the structures of materials are described.[b]Keywords[/b]: image analysis, surface layer, artificial intelligence, fuzzy logi