摘要
为提高三维模型的检索性能,将聚类分析用于特征描述符的提取以及模型间相似性关系划分等方面,能够对三维模型进行较为合理的分类,对较大规模三维模型数据库的索引和组织进行完善,提高三维模型检索效率。针对当前主流的基于聚类的三维模型检索算法进行分析,比较几种聚类算法的优势与不足,在其基础上进行改进,并继续应用于三维模型的检索中。
To improve the retrieval performance characteristics of the 3D descriptor, cluster analysis is used to extract shape descriptor, to divide similarity between models and some other fields, which could get a reasonable classification of 3D model, im- prove the index and organization of large-scale 3D model database, and enhance the efficiency of retrieval. The paper analyses the current mainstream model retrieval based on clustering, and compares advantages and disadvantages of several clustering algorithm to improve on these basis and apply to 3D model retrieval.
出处
《微型机与应用》
2012年第10期1-3,共3页
Microcomputer & Its Applications
