307 / 2023-10-12 18:33:56

A 3D Automated Modelling Method for Road Infrastructure

Vehicle-mounted mobile measurement,High-precision map,3D road modeling,Feature structure data,Delaunay triangulation

With the continuous growth in demand for intelligent transportation infrastructure development in China, there is widespread attention on how to rapidly achieve the three-dimensional digital representation of large-scale transportation infrastructure. In addressing issues such as low model accuracy, limited detail levels, poor adaptability, and mediocre suitability for different road alignment in road infrastructure three-dimensional modeling, this study focuses on a road 3D rapid modeling method with the road laser point cloud data obtained by vehicle-mounted mobile measurement system as the research object. Firstly, road scene objects are categorized, and high-precision vector data are extracted from road laser point cloud data, classifying them into planar, linear, and point features, which are then stored as spatial vector maps. Next, considering the model structures of various objects in road scenes, model feature data in standard formats are established based on feature data and modeling experience, along with the design of feature structure parameterized outputs to provide standardized data for rapid modeling of road scene elements. Subsequently, utilizing a 3D modeling platform, various key modeling methods are designed, including: 3D modeling method for road considering road alignment, automatic layout modeling of facility feature structure lines as 3D reference lines, and instantiation modeling of point facilities constrained by actual location and posture. Finally, experiments are conducted using a segment of vehicle-mounted laser point cloud data to extract feature structure data for various objects in road scenes, establishing a 3D modeling of road infrastructure. Evaluation is performed in terms of efficiency and accuracy, confirming the feasibility and precision of this method.