Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
8048652 | Manufacturing Letters | 2018 | 5 Pages |
Abstract
Shape density-based approaches have been extensively studied for 3D model matching due to its simplicity and explicit geometry descriptions. While most studies are concerned with either the matching capabilities or the theoretical aspects of the algorithms, this study experimentally investigates the descriptors' affine transformation invariance and noise robustness, as well as bin size effect. Six 3D density-based shape descriptors are derived and implemented. It has been found that the shape density descriptors depend on the bin size, all the descriptors are invariant to affine transformations, and all except Convex Hull are robust to noise.
Related Topics
Physical Sciences and Engineering
Engineering
Control and Systems Engineering
Authors
Xin Lin, Kunpeng Zhu, Wen-Feng Lu,