Evaluation of the accuracy of 3D face image capture using budget mobile applications

Abstract

Facial anthropometry as a diagnostic and analytical tool has been increasingly popular in orthodontics. This project aims to find an alternative to the currently available high-cost imaging equipment that satisfies a strictly high standard of accuracy in producing a fully textured facial scan. A set of 12 mannequin heads was anatomically landmarked and scanned using a verified industrial-level 3D scanner, followed by two budget mobile phone applications. The comparison includes 3D model surface analysis and assessment of the accuracy of image texture warping. The results showed a comparable level of accuracy of one of the applications (EM3D) which showed a close 3D model topographical match with the model created by the industrial scanner (mean square root distance (0.175 mm ± 0.104). In addition to that, there was a low level of error average (0.243 ± 0.274 mm) associated with 3D image texture warping. The results of this project showed that mobile phone applications can be considered as alternative diagnostic and assessment tools in clinical and research environments.