一种基于拓扑连接图的三维模型检索方法

3D Model Retrieval Based Topology Connection Graph

提出了一种基于分割技术的拓扑连接图构造方法 ,并用于三维模型检索 .首先构造多层邻域计算用于度量每个三角片平坦度的信号值 ,根据这些信号值采用一种改进的分水岭方法把模型分割成不同的曲面片 .然后利用这些曲面片之间的连接关系构造拓扑连接图 .最后通过比较不同模型的拓扑连接图相似性来得到它们的匹配度 ,根据匹配度进行三维模型检索 .

Content-based 3D model retrieval is gradually emerging as one of the most active research topics. In this paper, a 3D retrieval method based on topology connection graph (TCG) is proposed. The method decomposes a 3D model into some meaningful patches, and defines TCG from the connection relationships between patches. To achieve robust mesh segmentation for TCG construction, flatness measure for mesh faces is defined and used for mesh segmentation. A two-stage merging strategy is presented in the paper to avoid over-segmentation, a problem often occur in traditional segmentation algorithms. From the segmentation, patches of the 3D model are extracted, and related attributes of these patches are defined. The TCG of the 3D model is then constructed. The similarity between 3D models is computed by comparing their TCGs, and is used to match 3D models in the model database searching. Experiments show that, compared with similar algorithms, the method proposed in this paper can achieve better retrieval precision and efficiency.