Generative LOD algorithm based on space–time grid model

Abstract

The digital twin concept, which originated in the aerospace and industrial fields, has taken root in various industries such as construction. In terms of city management, digital twin city establishes a real-time interactive digital city by digitizing all the elements of the physical world, which promotes collaborative and intelligent urban planning, management and operation, and greatly improves the efficiency and accuracy of urban decision-making. However, large-scale city information model (CIM) often has a complex geometric structure and huge data volume, which brings great challenges to data storage, computation and visualization. Based on this, this paper proposes generative levels of detail (LOD) algorithm based on space–time grid model. Taking the space–time grid model as the space–time guiding framework, different levels of regions are divided according to the distance and viewpoint of the camera, and then the culling operation is performed, and then the appropriate level of detail and corresponding sampling method are dynamically selected for rendering to achieve lightweighting of the CIM. Finally, through the space–time grid visualization experiments and LOD experiments, it is proved that the method of this paper can significantly improve the visualization and loading efficiency of CIM.