Abstract: 3D shape retrieval based on content is short of local description for 3D shape. Aiming at this problem, this paper proposes an algorithm based on POint Description Image(PODI). On the 3D shape, it chooses a set of points from the 3D shape within some rule, calculates PODI of local shape distribution, gets a POint Description Image Set(PODISet) of PODI. PODISet is as the 3D shape descriptor. Experimental result shows that PODISet has better performance of local shape distribution compared with other two kinds of algorithm.

Key words: 3D shape retrieval, shape distribution, POint Description Image(PODI), POint Description Image Set(PODISet)

摘要： 基于内容的三维形状检索缺少对三维形状的局部描述。针对该问题，提出一种基于点描述图像的算法。在三维形状上，根据对应规则选取一些点，利用这些点计算局部形状分布的点描述图像(PODI)，得到PODI的点描述图像集合(PODISet)，将PODISet作为三维形状的描述符进行检索。实验结果表明，相比其他2种算法，PODISet的局部形状分布性能较优。

关键词: 三维形状检索, 形状分布, 点描述图像, 点描述图像集合

CLC Number: 

• TP393