How to automatically generate a realistic large-scale 3D road network is a key point for immersive and credible traffic simulations. Existing methods cannot automatically generate various kinds of intersections in 3D space based on GIS data. In this paper, we propose a method to generate complex and large-scale 3D road networks automatically with the open source GIS data, including satellite imagery, elevation data and two-dimensional(2D) road center axis data, as input. We first introduce a semantic structure of road network to obtain high-detailed and well-formed networks in a 3D scene. We then generate 2D shapes and topological data of the road network according to the semantic structure and 2D road center axis data. At last, we segment the elevation data and generate the surface of the 3D road network according to the 2D semantic data and satellite imagery data. Results show that our method does well in the generation of various types of intersections and the high-detailed features of roads. The traffic semantic structure, which must be provided in traffic simulation, can also be generated automatically according to our method.
翻译:如何自动生成现实的大型 3D 道路网络是沉浸和可信的交通模拟的一个关键点。 现有方法无法根据地理信息系统数据自动生成3D空间的各种交叉点。 在本文中,我们提出一种方法,用开放源GIS数据自动生成复杂和大规模 3D 道路网络,包括作为输入的卫星图像、高程数据和二维(2D) 道路中心轴数据。 我们首先引入道路网络的语义结构,以便在3D 场景中获取高度详细和完善的网络。 然后,我们根据语义结构和2D 道路中心轴数据生成2D 形状和地形数据。 最后,我们根据 2D 语义数据和卫星图像数据对3D 道路网络的表面进行分解和生成。 结果表明,我们的方法在生成各种交叉点和道路高偏差特征方面效果良好。 交通语义结构必须由交通模拟提供,也可以根据我们的方法自动生成。