基于距离和曲率特征的三维模型检索研究  

Study on 3D Model Retrieval Based on Distance and Curvature Feature

作　　者：张瑞[1] 方宗德[1] 李爱民[1] 张军辉[1] 

机构地区：[1]西北工业大学机电学院,西安710072

出　　处：《计算机工程》2010年第2期269-271,共3页Computer Engineering

摘　　要：针对采用单一特征描述基于内容的三维模型检索时检索精度不高的问题,结合全局描述的距离特征和局部描述的曲率特征,提出一种新的三维模型检索算法,通过距离特征相似性和曲率特征相似性的加权和,确立模型新的相似度匹配。仿真实验结果表明,相对于单一形状特征的模型检索,该检索算法能够获得更准确的检索效果。Aiming at the problem of low retrieval accuracy when single feature describes content-based 3D model,the distance feature of global description is combined with curvature feature of local description and a novel 3D model retrieval algorithm is proposed. The similarity match is identified by using the weight sum of similarity of both distance feature and curvature feature. Simulation experimental results show that,compared with the model retrieval of single shape feature,this algorithm can obtain better retrieval effect.

关 键 词：距离特征 曲率特征 三维模型检索 特征描述 

分 类 号：N945.12[自然科学总论—系统科学]