2D-PCA Shape Models: Application to 3D Reconstruction of the Human Teeth from a Single Image

• Published in: Medical Computer Vision. Large Data in Medical Imaging pp. 44 - 52
• Main Authors: Abdelrehim, Aly S., Farag, Aly A., Shalaby, Ahmed M., El-Melegy, Moumen T.
• Format: Book Chapter
• Language: English; Japanese
• Published: Cham Springer International Publishing 2014
• Series: Lecture Notes in Computer Science
• Subjects: Human Tooth; Perspective Camera; Reflectance Model; Tooth Shape; Tooth Surface
• ISBN: 9783319055299; 3319055291
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-05530-5_5

Accurate modeling of human teeth is mandatory for many reasons. Some of them are (1) Providing comfort to patients during mold process, and (2) Enhancing the accuracy level for oral orthodontist and dental care personnel. Shape from Shading (SFS) proved to be of great help in this problem, considering the fact that human tooth surface is textureless. In this paper, we address the problem of 3D tooth reconstruction, and improve the current algorithms. The approached developed in this paper reconstructs the teeth from single image shading with 2D-PCA shape priors that have more sophisticated reflectance model. Oren-Nayar-Wolff model was used for modeling the surface reflectance. This formulation uses shape priors as got from a set of training CT scans of real human teeth. Our experiments show promising quantitative results, which builds the infrastructure for having an optical based approach that accounts for inexpensive and radiationless human tooth reconstruction.