Evaluation of the accuracy of a mobile and a stationary system for three-dimensional facial scanning.

Purpose: Numerous three-dimensional (3D) facial scanners have emerged on the market; however, publications evaluating their accuracies are sparse. In this study, the accuracy of two 3D scanners used in facial scanning was evaluated.

Materials And Methods: A test specimen was attached at the right cheek and the forehead of 41 volunteers. These volunteers were scanned with Artec EVA and FaceScan3D. The acquired data were aligned to a 3D model of the test specimen for comparing the mean error, original length and width and angles to the measured values.

Results: The mean error in Best Fit alignment is significantly lower using Artec EVA (p < 0.001) for both test specimens. The deviation from the original length and width is significantly lower for the test specimens (p < 0.01) when measured with Artec EVA. The aberration of the angles measured between the front plane and the side plane is significantly lower when measured with Artec EVA (p < 0.001). Captured with Artec EVA the discrepancy between the original angle and the angle measured between the side planes to each other is significantly lower (p < 0.01).

Conclusions: Scanning with Artec EVA leads to more accurate 3D models as compared to scanning with FaceScan3D. The exactness achieved by both scanners is comparable to other scanners mentioned in literature.