3 D model reconstruction using ground laser scanning and drone imagery
In response to the challenges of obtaining accurate three-dimensional information from drone imagery in complex environments,such as densely vegetated ground areas and city centers with tall buildings,a method for high-quality three-dimensional model reconstruction in such complex areas is proposed using a combination of ground laser scanning and drone imagery technology.Imagery information and point cloud data of Building No.5 at Henan University of Urban Construction were obtained using drones and ground laser scanners.By em-ploying a single-point registration algorithm,named feature points were identified and matched between the ground laser point cloud and the drone imagery point cloud,forming a complete point cloud dataset.The ICP algorithm optimized by bidirectional KD-tree was utilized to reconstruct the three-dimensional realistic model of Building No.5,and its accuracy was analyzed both quantitatively and qualitatively.The results demonstrate that the three-dimensional model obtained using the integrated laser scanning and drone imagery technology is more detailed compared to that obtained from drone imagery alone,accurately reflecting the terrain and building structures of the study area.The model′s point position mean error is 0.039 m,achieving a centimeter-level precision that meets the accuracy requirements for practical applications.
ground laser scanningdrone imagery3D model reconstructiondata fusion