Mapping of Color Information from Camera Images to 3D Models for the Manufacturing of Aesthetic Dentures.

Abstract

Aesthetic dentures are created to permanently replace missing teeth in a way that is as close to the patient's natural dental condition as possible. Recreating a natural color impression, especially for vestibular faces, is a major aspect in this process. Colors vary greatly from patient to patient and so do the shades of dental materials provided by professional suppliers. Until now, the manufacturing process of these highly customized artifacts is mostly carried out by hand by a trained technician who makes use of their work experience. However, digitization and automation find their way into dentistry. This paper introduces an algorithm to visualize the final result of a denture when a camera picture and 3D model is at hand. Several steps of image processing, such as segmentation, feature extraction and inpainting for black spots and reflections, are carried out. Subsequently, a projection method to map the refined image onto the 3D model is discussed. In addition, rendering of the so-colored models is mentioned as a potential extension. Implementation, which has been realized in form of C++ code using freeware frameworks OpenCV and CHAI3D, is referred to in the course of explaining the respective steps. Finally, results are presented and an outlook on future work is given.