Abstract: A method for categorizing a fracture surface of a component. A digital image of the fracture surface is generated. The digital image or one or more segments of the digital image is/are analyzed, wherein a category of a fracture pattern is assigned to one or more regions of the digital image or of one or more segments based on detection and statistical evaluation of spatial distribution of the smallest homogeneous image units (pixels) or image elements, and, provided that multiple regions and/or multiple segments of the digital image were analyzed, information about the regions and/or details of the digital image and categories of fracture patterns determined and assigned to them are linked to each other, and a category is assigned to the fracture surface based on this link. The method permits automatic, self-learning-categorization of fracture surfaces, for example for series tests carried out for joining technologies.

Abstract: A method for categorizing a fracture surface of a component. A digital image of the fracture surface is generated. The digital image or one or more segments of the digital image is/are analyzed, wherein a category of a fracture pattern is assigned to one or more regions of the digital image or of one or more segments based on detection and statistical evaluation of spatial distribution of the smallest homogeneous image units (pixels) or image elements, and, provided that multiple regions and/or multiple segments of the digital image were analyzed, information about the regions and/or details of the digital image and categories of fracture patterns determined and assigned to them are linked to each other, and a category is assigned to the fracture surface based on this link. The method permits automatic, self-learning-categorization of fracture surfaces, for example for series tests carried out for joining technologies.