Abstract: The authors propose a new 3D shape representation, “local bag-of-words models”and use overlapped “local bag-of-words models”which improves the robustness of the algorithm by avoiding shape segmentation. The detailed steps are also proposed to apply the “Earth Mover’s distance”and “pyramid match kernels”algorithms to quantify partial shape similarity. It proves that the proposed distance measure mitigates the distance under-estimation problemof global bag-of-words model. Finally, the experiments firmly demonstratethe effectiveness of the proposed approach to 3Dpartial shape retrieval.

Key words: 3Dshaperetrieval, Earth Mover’s distance, pyramid match kernels, partial similarity measure, bag-of-words model

摘要： 提出一种新的基于“局部字袋混合模型”表示的三维部分形状检索方法。该方法首先定义一种支撑范围相互重叠的三维局部字袋模型表示, 并在该形状表示下, 给出“推土机距离”的基础距离度量的定义, 并运用“层次匹配核”方法计算三维模型间部分相似性的具体步骤, 显著提高了原方法的效率和鲁棒性。在理论上证明了所提出的距离度量改进了整体字袋模型低估三维形状之间距离的问题。实验结果显示了该方法对三维部分形状检索的有效性。

关键词: 三维形状检索, 推土机距离, 层次匹配核, 部分相似性度量, 字袋模型