三维模型库语义网构建及检索方法研究  

Research on construction of semantic web of 3D models database and retrieval method

作　　者：贾晖[1] 刘建元[1] 张建刚[2] 

机构地区：[1]西安邮电学院计算机学院,陕西西安710121 [2]西安热工研究院有限公司电站信息及监控技术部,陕西西安710032

出　　处：《西安邮电学院学报》2012年第3期53-57,共5页Journal of Xi'an Institute of Posts and Telecommunications

摘　　要：为了实现三维模型语义检索,描述了一种基于本体来构建三维模型库语义网的方法,并在此基础上实现基于语义的三维模型检索。该方法首先建立三维模型库本体描述,根据模型库中的内容提取类,对象和属性。其次利用Word-Net英文本体查找本体原始结点的语义扩展结点,包括同义词,上位词,下位词和整体部分关系词,从而构建语义网络。在语义网的基础上实现基于SPARQL的三维模型检索。结果表明,该方法能有效地扩充一个分类粒度较粗的三维模型库的语义内容,扩大语义覆盖面,同时扩充词汇之间的语义相关性,从而提高基于关键词语义检索的准确性。In order to realize semantic retrieval of 3D models, a new method to construct semantic web of 3D model database based on ontology is presented, and the 3D models retrieval based on semantic web is realized. Firstly ontology of three-dimension model database is built, and then it extracted classes, objects and attributes according to the model. Secondly, WordNet which is an English ontology is utilized to expand original ontology node to semantic extension node, including synonym.hypernym.hyponym and holonym, so as to construct semantic web. On the basis of semantic web and based on SPARQL,it realized the 3D model retrieval. Experiment result shows that this method not only effectively expands the semantic vocabularies of a 3D model database, but also keeps good semantic relevance of the expanded vocabularies to the original ones, so as to achieve semantic based 3D model retrieval effectively, thus, the accuracy of semantic retrieval based on keywords improved.

关 键 词：三维模型检索技术 三维模型库 WORDNET 语义网 

分 类 号：TP37[自动化与计算机技术—计算机系统结构]