**5. Conclusions**

In this paper, models based on machine learning were developed to predict surface quality from photographs of Ti6Al4V parts created by single point incremental forming. Training data were collected using a surface profilometer and a stereoscopic microscope,

and more than one hundred and fifty simulations were run to optimize the processing parameters and to find the most suitable classifier. The main findings derived from the current work are as follows:


Overcoming the limitations of this work will provide the basis for the use of a regression model, instead of a classification one, to predict continuous values. One future direction will be to build a large and uniform database to avoid a class imbalance and to apply normalization and filtering techniques in order to obtain stable roughness values and to reduce the noise in pictures, respectively. In this way, it will be possible to build a reliable model with high detection capacity to be deployed in a productive environment.

In the future, this measuring strategy is expected to be used on real-time video transmission in a manner that enables automatic and fast selection of parts according to tolerance requirements.

**Author Contributions:** V.M. and F.G.-S.: Conceptualization. V.M., F.G.-S., F.B.-M. and J.N.: Data curation. V.M., F.G.-S. and F.B.-M.: Formal analysis. F.G.-S. and V.M.: Funding acquisition. V.M., F.G.-S. and J.N.: Investigation. F.G.-S., V.M. and F.B: Methodology. V.M.: Project administration. V.M.: Resources. F.G.-S. and F.B.-M.: Software. V.M., F.G.-S. and M.C.M.: Supervision. V.M. and F.G.-S.: Validation. F.B.-M., F.G.-S. and V.M.: Visualization. V.M., F.G.-S., F.B.-M., J.N. and M.C.M.: Writing—original draft. V.M., F.G.-S. and M.C.M.: Writing—review and editing. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**

