您的位置:首页  > 论文页面

基于语义相关性的三维模型分类及标注

发表时间:2009-10-31  浏览量:1712  下载量:611
全部作者: 王新颖,吕天阳,王钲旋
作者单位: 吉林大学计算机科学与技术学院;哈尔滨工程大学计算机科学与技术学院
摘 要: 提出一种基于语义相关性的三维模型半自动标注及分类方法。三维模型之间的语义相关性来自于用户在基于形状检索过程中的相关反馈记录,并可以通过少量已标注模型自动推断出未知模型的标注信息。对普林斯顿形状基准数据库(Princeton shape benchmark,PSB)的实验表明:该方法在三维模型的语义分类以及标注中取得了较好的效果。
关 键 词: 计算机应用;相关反馈;聚类;语义标注
Title: Classification and annotation of 3D models based on semantic relationship
Author: WANG Xinying, LV Tianyang, WANG Zhengxuan
Organization: College of Computer Science and Technology, Jilin University;College of Computer Science and Technology, Harbin Engineering University
Abstract: The paper explores the semi-automatic semantic annotation and classification based on semantic relationship. The semantic relationship among 3D models came from the uses’ retrieval records of shape-based retrieval and is adopted in the automatic semantic annotation of the models whose semantic is unknown. The experiments on princeton shape benchmark show that the proposed method achieves good performance not only in semantic classification but also in semantic annotation.
Key words: computer application; relevance feedback; clustering; semantic annotation
发表期数: 2009年10月第20期
引用格式: 王新颖,吕天阳,王钲旋. 基于语义相关性的三维模型分类及标注[J]. 中国科技论文在线精品论文,2009,2(20):2136-2142.
 
0 评论数 0
暂无评论
友情链接