三角网内插及约束的精细影像点云构建  

Fine image point cloud construction based on triangulation interpolation and constraint

作　　者：胡春梅 夏国芳 张旭 王伟 尉丽霞 HU Chunmei;XIA Guofang;ZHANG Xu;WANG Wei;WEI Lixia(School of Geomatics and Urban Spatial Informatics,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;China Cultural Heritage Information and Consulting Center,Beijing 100029,China;Shandong ZhengYuan Digital City Construction Co.,Ltd.,Yantai,Shandong 264670,China)

机构地区：[1]北京建筑大学测绘与城市空间信息学院,北京100044 [2]中国文物信息咨询中心,北京100029 [3]山东正元数字城市建设有限公司,山东烟台264670

出　　处：《测绘科学》2022年第4期87-93,168,共8页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41401536);北京建筑大学市属高校基本科研业务费专项资金资助(X18230)。

摘　　要：针对目前几何及纹理复杂对象影像建模在点密度方面不能按需获取及边缘精度不高等问题,该文提出了基于均匀同名特征三角网内插及约束的精细点云构建方法。首先以格网形式提取并匹配同名均匀特征点对,构建同名三角网;其次在上述三角网约束下,以三角形重心为条件,以面积作为精细程度的阈值,确定内插重心同名点,同时,对边缘也进行特征提取和高精度匹配;最后应用前方交会方法生成具有需求点密度和高精度边缘特征的三维影像点云。实验证明,该文方法在点云密度和边缘信息上都有一定的优势,提高了对象点云模型的整体精度。In order to solve the problem that the point density can not be obtained on demand and the edge accuracy is not high in the current image modeling of complex geometric and texture objects,this paper proposed a fine point cloud construction method based on uniform homonymic feature triangulation interpolation and constraint.Firstly,the uniform feature points of the same name were extracted and matched in the form of grid,and the triangulation of the same name was constructed;secondly,under the constraints of the triangular network,the homonymous point of the interpolation center of gravity was determined with the center of gravity of the triangle as the condition and the area as the threshold of fineness;At the same time,feature extraction and high-precision matching were also carried out for the edge;finally,the forward intersection method was used to generate the 3Dimage point cloud with the required point density and high-precision edge features.Experiments showed that this method had certain advantages in point cloud density and edge information,and improved the overall accuracy of the object point cloud model.

关 键 词：三角网 内插 约束 点密度 边缘 

分 类 号：P237[天文地球—摄影测量与遥感]