Refined 3D road modeling method combined with road alignment

ObjectiveThe existing 3D road modeling methods often suffer from low accuracy, insufficient detail, and poor adaptability, failing to meet road maintenance requirements. To address these issues, this study proposes an automated 3D modeling algorithm that integrates road alignment information. MethodFirst, starting from the data characteristics of mobile laser point cloud combined with road alignment, the road centerline constraints were introduced. Alpha Shapes algorithm was employed to preprocess data for extracting road boundary points and pavement feature points. Next, an ordered point cloud boundary constraint algorithm was developed based on incremental Delaunay triangulation. This algorithm identified and removed triangles outside the pavement domain, adapting to various road alignments and enhancing reconstruction accuracy. Finally, the meshing effect, quality, and accuracy of the generated 3D model were evaluated to verify the feasibility of algorithm. ResultIntegrating road alignment constraints during feature point extraction yields a simplified point cloud that retains key pavement characteristics. It significantly reduces data complexity and processing time. Furthermore, the proposed boundary constraint algorithm effectively eliminates extraneous triangles during triangulation. It ensures the model aligns perfectly with the actual road contours, satisfying modeling requirements for diverse road alignments. ConclusionThe proposed algorithm significantly improves the precision and adaptability of 3D pavement reconstruction. By overcoming the limitations of existing methods regarding accuracy, detail level, and adaptability, it effectively meets the practical demands of road management and maintenance.