| Summary: | 碩士 === 國立中正大學 === 機械工程學系暨研究所 === 99 === This research proposes a new approach to reconstruct a complete tooth model by integrating information from Computer Tomography(CT) images and scanned crown data. First, the reference model extracted from CT image using Marching-Cube algorithm could be applied as a target. Then, the plaster model could be digitized by laser scanner and processed to get the crown data with higher precision. At the same time, the Level set method is used to segment individual root contours from CT images precisely. The crown data and roots contours information could be integrated by Iterative Closest Point(ICP) algorithm. Finally, the integrated model as an unorganized point clouds and reconstructed as 3D surface model by Delaunay triangulation algorithm.
Compared to other related works, our tooth root contours are segmented out along long axis from the CT image sequences. The Level set method can separate adjacent touching teeth correctly and generate the virtual boundary individually. The integration of scanned crown and extracted tooth roots could be reconstructed as a high precision tooth model. The result finally could be applied to orthodontic treatment planning system to demonstrate the performance of the proposed method.
|
|---|
