Validation of a High-Resolution 3D Face Scanner Based on Stereophotogrammetry

Emerging 3D surface acquisition technologies have recently introduced new face scanning methodologies. Among passive methods, stereophotogrammetry has proved to be particularly promising in the evaluation of facial morphology. The aim of the current study is to propose and study a new and structured validation method for 3D face scanning system based on close-range stereophotogrammetry. The authors investigated accuracy of the FaceshapeMaxi5 3D photogrammetric scanner developed by Polishape 3D srl (spin-off of Politecnico di Bari). Measurements were taken over a set of 23 anthropological soft-tissue facial landmarks marked on two different dummies. The validation procedure consisted in: 1) determining the most appropriate photo-shooting parameters; 2) assessing the photogrammetric software precision; 3) testing the system measurements against those of a CMM. Critical was the camera system calibration process. A new method was implemented, namely full-field calibration. Accuracy was evaluated using different indicators. The operator error was measured by repeatedly digitising landmarks on the 3D model and it was within 0.059mm. The reproducibility error was calculated by digitising landmarks on two different occasions. The Euclidean distance between the two matched sets of coordinates was thus computed resulting within 0.090mm. Each dummy was digitised using a CMM of documented accuracy (0.5μm). The so obtained landmark coordinates were considered as the “gold standard”. After Procrustes alignment, the displacement between the adjusted system coordinates and the “gold standard” was calculated. Therefore, the achieved results proved that the presented 3D face scanner was reliable enough for capturing facial morphology for clinical treatment and anthropological practice. The validation method is applicable to every kind of 3D Facial Scanner, requiring uniquely the presence of the textured and tessellated point cloud.