摘要
针对基于内容的三维模型自动分类问题,提出一种密度峰值驱动的三维模型无监督分类算法.首先利用多种特征描述符分别对每个三维模型提取相应的特征向量;然后将得到的特征向量运用鲁棒主成分分析去除噪声并降维;最后通过计算特征向量分布的密度峰值,并配合决策图,以直观的方式确定三维模型分类类别数,最终实现三维模型的无监督分类.实验结果表明,与传统算法相比,该算法具有易于确定分类类别数、准确率高、鲁棒性强等优点.
In this paper, we propose an unsupervised classification algorithm by using density peaks for automaticcontent-based 3D model classification. Firstly, the algorithm extracts multiple kinds of feature vectors for eachmodel in the given shape collection. Secondly, it uses robust principal component analysis to denoise the featurevectors and reduce their dimensions simultaneously. Finally, the algorithm determines the number of categories ofthe 3D models and realizes an unsupervised classification in an intuitive and visual way by computing the densitypeaks of the feature vectors’ distribution and a corresponding decision graph. Extensive experimental results showthat the number of categories of clustering is much easier to determine and the results are more accurate and robustin our algorithm when compared with the traditional algorithms.
作者
舒振宇
祁成武
辛士庆
胡超
韩祥兰
刘利刚
Shu Zhenyu Qi Chengwu Xin Shiqing Hu Chao Han Xianglan Liu Ligang(School of Information Science and Engineering, Ningbo Institute of Technology, Zhejiang University, Ningbo 315100 School of Electronic Information Engineering, Taiyuan University of Science and Technology, Taiyuan 030024 Faculty of Electrical Engineering and Computer Science, Ningbo University, Ningbo 315211 School of Mathematical Sciences, University of Science and Technology of China, Hefei 230026)
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2016年第12期2142-2150,共9页
Journal of Computer-Aided Design & Computer Graphics
基金
国家自然科学基金(11226328
61222206
61300168
61273332)
浙江省自然科学基金(LY17F020018)
宁波市自然科学基金(2012A610018)
宁波市创新团队资助项目(2014B82015)
关键词
三维模型
分类
密度峰值聚类
鲁棒主成分分析
3D model
classification
density peak clustering
robust principal component analysis
