中文摘要：

为了解决基于文本和基于内容的三维模型检索所存在的问题,提出语义树在三维模型检索中的应用,该方法首先基于WordNet为PSB（Prince.tonShapeBenchmark）本体库构建三维模型语义树,提出一种全新的构造语义树的方法,而且扩充语义树也十分简单,在进行检索时采用信息量来计算语义相似度返回相似性高的模型,并且提出了较好的检索方法,同时通过反馈机制进一步优化检索结果,实验表明该方法可以有效地提高三维模型检索的效率。

英文摘要：

In order to solve the problem based on the text and 3D model retrieval,the application of semantic tree in 3D model retrieval is put forward,The method firstly builds 3D models by semantic trees of PSB（Prince.tonShapeBenchmark）ontology library based on WordNet.The method presents a new structure of semantic tree,and extended semantic tree is also very simple,and then in the retrieval the calculation of semantic similarity is used to return to high similarity model,and a better retrieval method is put forward,through a feedback mechanism to further optimize the retrieval results,The experiments show that the method can effectively improve the efficiency of 3D model retrieval.