A digital image analysis method is presented for the identification and quantification of different graphite forms in spheroidal graphite iron (SGI) castings. SGI castings with low nodularity as well as with graphite degenerations, such as chunky graphite, typically exhibit lower mechanical properties and often do not meet minimum tensile strength requirements. Therefore, a fast and accurate digital image analysis method for the quantification of the graphite morphology is important for the material production and quality control. In the presented method, the parameters for the classification of the graphite particles are determined on the basis of the form classes and their corresponding reference image in the standard ISO 945-1. Further parameters are defined for the identification of graphite degenerations, such as chunky graphite. Based on the obtained micrographs and their analyzed graphite morphology, especially with focus on chunky graphite, a machine learning algorithm is trained. This algorithm can be applied for an automated digital image analysis method and used as a preliminary examination. The machine learning model informs about the occurrence of graphite degenerations and the necessity of more detailed micrographs and examinations due to the presence of graphite degenerations. Together with the implemented parameters for the graphite form classes and degenerations in case of the more detailed examinations, the automated digital image analysis method allows a fast and accurate determination of the graphite morphology in foundries' quality assurance.