In recent years, Internet of Things has being used in several fields of modern life. The possibility of having an interconnection between all the related objects through data opens up a world of research in different fields. Among these fields, IoT technology is considered to make a significant impact in casting industry and the whole casting process. Moreover, the casting industry is an exceptionally critical field since it directly provides elements to other industries. Therefore, castings are subjected to a series of very rigorous quality controls which must be validated by the entire manufacturing process. Considering that the casting process has to be interconnected in order to foresee the state of the each piece before it is produced, it is vitally important that all sub-processes feed back into the system to learn from errors. In this work, we presented a new methodology for the detection and categorization of the imperfections on the surface of the castings. Among these defects, we particularly focused on inclusions, cold laps and misruns. To this end, we first compared several features extracted from the obtained images in order to highlight the regions of the casting that may be affected. And then, we applied several machine-learning techniques to classify the regions. The final results were carried to the starting point of the process to use them in the models to predict the quality of new castings.