In recent years the affirmation of industry 4.0 has marked a trending towards new means of production and more sophisticated control infrastructures. An example of this trend is the widespread application of additive manufacturing machine, alongside innovative software and control techniques, as testified by the application of Machine Learning in production processes and the shift towards an IoT approach. This thesis project is part of the application of a Machine Learning algorithm to a Selective Laser Melting (additive manufacturing technique) machine in order to optimize the machine parameter to prevent defect formation. In particular, the part covered by this thesis is the feature extraction from the process: using a photo camera, pictures of the building process in a Selective Laser Melting machine are taken, then such pictures are analyzed through Python methods to extract relevant features that will be later fed to the optimization and control part..