基于形状分布的三维骨骼的相似性度量

Measuring 3D-Skeleton Similarity Based on Shape Distribution

文中应用形状分布算法度量由CT数据重建出的三维骨骼的相似度,并提出了改进。基本步骤是:在模型表面选择随机点,计算每两点间的距离,以形状分布函数来构建形状分布直方图,通过比较形状分布直方图来给出不同骨骼间的相似度。实验表明,该方法能够准确度量骨骼的相似性。

Uses shape distribution to measure similarity of 3D skelecton which is reconstructed from CT data in this paper. The approach of shape distribution consists of these steps: select the number of random point on the model surface; compute the distance of each point pair; construct the shape distribution histogram by using shape function; measure similarity of 3D models by comparing shape function. The experimental result demonstrated the benefits of our method in precision and computational complication.