点云数据与倾斜影像的古建筑三维模型构建

3D model construction of ancient buildings based on point cloud data and oblique images

针对古建筑三维建模中使用单一技术手段测量导致的结构缺失、精度低、图像拉花和破洞等问题,研究基于倾斜影像与点云数据的古建筑三维模型构建方法。使用无人机和激光扫描仪采集古建筑的倾斜影像数据和激光点云数据,并预处理上述数据,通过新特征点配准算法实现两种数据融合,得出准确完整的古建筑三维数据。提取古建筑的轮廓线,使用Revit软件构建古建筑三维模型,对所得模型纹理贴图,实现古建筑的三维模型构建。实验结果表明:该方法可以获取完整的、高精度的古建筑三维数据,以此构建的古建筑三维模型的精准度和完整度都很高,而且结构清晰、纹理清楚、效果逼真,具有优秀的建模性能。

The problems of structure loss,low accuracy,image distortion,and holes caused by the use of single technical means in three-dimensional(3D)modeling of ancient buildings occur during measurement.Therefore,a method for constructing 3D models of ancient buildings based on oblique images and point cloud data was studied.Unmanned aerial vehicles(UAV)and laser scanners were utilized to collect oblique image data and laser point cloud data of ancient buildings,and the above data was preprocessed.The two kinds of data were fused through a new feature point registration algorithm,so as to obtain accurate and complete 3D data of ancient buildings.The contour lines of ancient buildings were extracted,and Revit software was used to construct 3D models of ancient buildings.The resulting model texture was mapped to achieve 3D model construction of ancient buildings.The experimental results show that this method can obtain complete and highprecision 3D data of ancient buildings,and the accuracy and completeness of the 3D model of ancient buildings constructed based on this method are high.Moreover,the structure is clear;the texture is clear,and the effect is lifelike,indicating excellent modeling performance.