Bio-Information Based Segmentation of 3D Dental Models

The automatic and accurate segmentation of 3-D dental model digitized through intra or extra oral measurement methods is an important step in computer aided orthodontic system. This paper presents a robust segmentation method based on the bio-information of individual\´s teeth with little manual interaction. The discrete curvature of the dental model is first computed; using the computed discrete curvature as the reference, and then the feature regions in which the segmentation boundary may be included is filtered out according given curvature threshold. The feature regions including segmentation boundary are "bridged" interactively. The skeleton of the feature regions is extracted according to the bio- information of the dental model, and the segmentation boundary is got after branches removing. Finally, the teeth of the dental model are separated one by one along the segmentation boundary.