摘要
针对古村落对三维实景模型的高精度、精细化真实还原需求,该文设计了一种基于影像进行三维实景重建的技术方法。采用多旋翼无人机和手持三轴云台相机分别采集低空和地面的多视角影像,通过Context Capture软件对来自地面、低空的多视角影像分别进行空中三角测量和区块合并,实现了地面、低空多视角影像的集成,并进行三维实景建模。该方法应用于东莞市南社古村落的三维实景建模工作中,结果表明,该方法建模成果的几何精度和纹理贴图满足相关规范要求,适用于大型文化遗产的三维实景建模。该方法充分利用低空和地面不同位置的视角的优势,能够为大型文化遗产采集较为完整的影像信息,在构建高精度、还原效果好的三维实景模型方面具有明显的优势。
Aiming at 3 Dreal remodeling for historical building groups in ancient villages with high precision and fine real reduction,a technology process based on images technology for 3 Dreal remodeling was put forward in this paper.During the image collecting,low altitude multi-view images were collected by multi-rotor unmanned aerial vehicle(UAV)and ground multi-view images were collected by three-axis handheld stabilized camera.In the technology process,firstly,images from low attitude and ground went through aero triangulation respectively and the productions of aero triangulation were merged into a same spatial block by Context Capture for high precision exterior orientation;secondly,based on the exterior orientation,Context Capture conduct 3 Dreal remodeling process,in which contained dense image matching,colored point cloud generating,point cloud optimization,triangulated irregular network(TIN)generating and texture mapping.This technology process applied in the 3 Dreal remodeling of Nanshe historical villages in Dongguan and the precision of production was evaluated by geometry measurement.In the 3 Dreal modeling results of large cultural heritage,preliminary experiments showed that the precision of geometry and the spatial resolution of texture mapping met relevant specification requirements.Therefore,this technology process made full use of the advantages of different position of the ground and low altitude,and it was able to collect more complete image information for large cultural heritage with obvious advantages in building a 3 Dreal scene model in high accuracy and good recovery.
作者
胡应龙
陈颖彪
郑子豪
吴志峰
李安敬
杨德校
HU Yinglong;CHEN Yingbiao;ZHENG Zihao;WU Zhifeng;LI Anjing;YANG Dexiao(School of Geographic Sciences, Guangzhou University, Guangzhou 510006, China;Guangdong Province Engineering Technology Research Center for Geographical Conditions Monitoring and Comprehensive Analysis, Guangzhou 510006, China;Guangzhou Yu'an Information Technology Co., Ltd.,Guangzhou 510663, China)
出处
《测绘科学》
CSCD
北大核心
2019年第4期139-145,共7页
Science of Surveying and Mapping
基金
广东省自然科学基金项目(2016A030313551)
广东省前沿与关键技术创新专项资金(省重大科技专项)项目(2017B010117004)
广东省国土资源科研专项(GDGTKJ2017001)
广州大学研究生创新研究资助计划项目(2017GDJC-M01)
