城市场景结构感知的网格模型简化算法  

Structure-aware simplified algorithm of mesh model for urban scene

作　　者：张春森[1] 张会 郭丙轩[2] 彭哲 ZHANG Chunsen;ZHANG Hui;GUO Binxuan;PENG Zhe(College of Geomatics, Xi’an University of Science and Technology, Xi’an 710054, China;State Key Laboratory of Information Engineering in Surveying, Mapping & Remote Sensing, Wuhan University, Wuhan 430079, China;Wuhan Xuntu Technology Co.Ltd.,Wuhan 430079, China)

机构地区：[1]西安科技大学测绘科学与技术学院,陕西西安710054 [2]武汉大学测绘遥感信息工程国家重点实验室,湖北武汉430079 [3]武汉讯图科技有限公司,湖北武汉430079

出　　处：《测绘学报》2020年第3期334-342,共9页Acta Geodaetica et Cartographica Sinica

基　　金：国家自然科学基金重大研究计划(91638203);陕西省自然科学基金(2018JM5103)~~

摘　　要：针对二次误差测度(QEM)网格简化算法全局几何特征信息损失严重的缺点,提出一种具有结构感知功能面向城市三维模型重建的网格简化算法。该算法顾及城市影像中平面结构特征,以代理平面为全局特征约束条件,使模型在简化过程中全局结构特征更多地被保持,以利于多层次细节模型(LOD技术)、网格优化提速等模型后续操作。以倾斜摄影获取影像生成的初始三角网格模型为试验数据,采用所给算法对其进行网格简化并与QEM算法进行对比。结果表明:所给算法简化精度及简化效率均优于QEM算法。Aiming at the shortcomings of global geometric feature loss caused by quadratic error measure(QEM)mesh simplification algorithm,a structure-aware mesh simplification algorithm for urban environment 3D model reconstruction is proposed.The algorithm takes into account the planar structure features in urban images,and uses the proxy plane as the global feature constraint in the simplification process,so that the global structural features of the model are more preserved during the simplification process,which is conducive to the refinement and progressive transmission of the model(LOD technology),grid optimization speedup and other model follow-up operations.In this paper,the initial manifold triangle mesh model generated by oblique photography is taken as the experimental data.The proposed algorithm is used to simplify the mesh and compare with the QEM algorithm,which experimental results show that the proposed algorithm has excellent simplification accuracy and simplified efficiency than the QEM algorithm.

关 键 词：网格简化 场景结构感知 三维重建 平面检测 

分 类 号：P231[天文地球—摄影测量与遥感]