The aim of a fractographic investigation is the evaluation of macroscopic and microscopic crack surface characteristics and, as a result, the determination of the fracture mechanism of a specimen or component. The basis for such evaluations of fracture characteristics comes from actual comparative mechanical testing and from the literature. Fractographic analysis can be complex and, in any case, requires considerable experience. In the presented work, software was developed that quantitatively determines fracture characteristics and fracture mechanisms utilizing digitized expert knowledge, deep neuronal networks, and standard 2D and topographical data from SEM imaging. Fracture surfaces with typical characteristics were obtained conducting fatigue testing experiments on various materials. After fatigue cracking the specimens were opened to generate forced fracture. Specimens were imaged using SE and BSE sensors of a SEM to generate comprehensive training data. Topographical data are obtained from 4Q-BSE detector using shape-from-shading technology. Using a combination of multiple input data, the most accurate classification of crack features was obtained. As a use case, the integration of the AI tool into the fractographic online database FractoDB is presented, and the benefit is exemplarily demonstrated.