基于模型内二面角分布直方图的非刚性三维模型检索  

Non-rigid 3D Shape Retrieval Based on Inner Dihedral Angle Histogram

作　　者：李海生[1,2] 孙莉[1,2] 吴晓群[1,2] 蔡强[1,2] 杜军平[3] Li Haisheng;Sun Li;Wu Xiaoqun;Cai Qiang;Du Junping(School of Computer and Information Engineering, Beijing Technology and Business University, Beijing 100048;Beijing Key Laboratory of Big Data Technology for Food Safety, Beijing 100048;School of Computer Science, Beijing University of Posts and Telecommunications, Beijing 100876)

机构地区：[1]北京工商大学计算机与信息工程学院,北京100048 [2]食品安全大数据技术北京市重点实验室,北京100048 [3]北京邮电大学计算机学院,北京100876

出　　处：《计算机辅助设计与图形学学报》2017年第6期1128-1134,共7页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(61320106006;61532006;61602015);北京市自然科学基金(4162019);北京工商大学两科基金培育项目(LKJJ2015-27)

摘　　要：针对非刚性三维模型检索中复杂曲面凹凸性特征和局部几何变化特征的提取问题,提出一种基于模型内二面角分布直方图的特征描述方法.首先对内二面角直方图统计特征进行了定义并对其性质进行探讨,给出特征提取的具体步骤;然后采用遗传算法进行多特征融合权重优化,提出基于融合特征的非刚性三维模型检索算法.在SHREC公布的非刚性数据集上进行实验的结果表明,内二面角分布直方图统计特征具有更强的区分能力和良好的算法效率,融合特征进一步提高了检索结果.This paper proposed a novel feature descriptor based on interior dihedral angle histogram for non-rigid3D shape retrieval to describe the concave-convex and local geometric features of3D models.We first definedthe interior dihedral angle histogram and analyzed its statistical properties.The specific steps were given to extractthe histogram feature.Then we employed the genetic algorithm to optimize the weights of various featuresand designed a non-rigid3D shape retrieval algorithm based on the combined feature.The experimental results onthe non-rigid dataset published by SHREC show that the interior dihedral angle histogram is more discriminativeand efficient.In addition the combined feature further improves the retrieval precision.

关 键 词：非刚性三维模型检索 内二面角分布直方图 遗传算法 融合特征 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]