基于多源数据融合古建筑三维重建技术的研究

Research on the 3D Reconstruction Technology of Ancient Buildings with Multi-source Data Fusion

将无人机倾斜影像、地面三维激光点云和数码相机照片相结合,从影像数据与点云数据2个层面研究数据融合的理论及策略,通过数据预处理、坐标系统一、数据配准、影像点云融合、点云数据融合实现三维实景重建。建立的三维实景模型,能够真实全面地还原和反映目标物全貌,且具备直观表达对象整体空间信息的能力。通过点、线、体3个层面的精度验证可以得出,本文提出的多源数据融合建模方法能够建立无遮挡、无空洞、细节表达完整的三维实景模型,且模型精度良好,为古建筑数字化保护拓展了一个新的方向。

This paper combines the drone oblique images,the ground three-dimensional laser point cloud and the digital camera photos.From the image data and the point cloud data,the theory and strategy of data fusion are studied.Through data preprocessing,coordinate system unification,data registration,image point cloud fusion,point cloud data fusion,it realizes 3D real scene reconstruction.The established three-dimensional real scene model can truly and comprehensively restore and reflect the full picture of the target,and has the ability to intuitively express the overall spatial information of the object.Through the accuracy verification at the three levels of point,line,and volume,it can be concluded that the multi-source data fusion modeling method proposed in this paper can establish a three-dimensional real-world model with no occlusion,no holes,and complete detail expression,and the model has good accuracy.It expands a new direction for the digital protection of buildings.