基于兼容性网格分割的骨骼参数计算

A New Measurement Method for the Morphological Parameters of Bone Models Based on Compatible Mesh Segmentation

摘要: 为了解决骨骼参数难以测量的问题，提出一种利用网格分割结果计算参数的方法.首先对骨骼网格模型在提取少量特征点的基础上进行语义分割；然后将分割得到的模板借助基于Laplacian变形的非刚性配准，指导同类骨骼模型实现快速兼容性分割；最后结合网格分割结果及特征点计算出骨骼的主要形态参数.实验结果表明，该方法仅需少量的用户交互就能获得较好的测量效果，结合模板的策略也支持大批量骨骼模型的参数测定.

Abstract: To measure the morphological parameters of 3D bone models effectively,a method based on mesh segmentation is proposed.The approach starts with a semantic-oriented segmentation of bone mesh using a few feature points.Then the segmented mesh acts as a template,guiding other same kinds of bone meshes to achieve similar decompositions through a non-rigid shape registration based on Laplacian editing.After mesh segmentation,parameters of each bone model are calculated combining with the feature points.Experimental results show that good measurement can be achieved even with small interaction.In addition,the measurement of the morphological parameters of numerous bone models are well supported by the template-related strategy.