A Realtime Rendering Framework of Large Dataset Environment Based on Precomputed HLOD

Abstract

Realtime rendering of large dataset environment including a lot of huge dataset models can be challenging. Discrete LOD and continuous LOD are the traditional methods for preprocessing model data of large dataset environment. Division granularity of discrete LOD is coarse while continuous LOD is complex and time-consuming. A realtime rendering framework of large dataset environment based on precomputed HLOD is proposed in this paper. As different granularity and hierarchy is used to divide the model and environment, the details can be improved and the time will be shorter while rendering. A format of HLOD file is proposed as the time for HLOD preprocessing of models can be long, result of preprocessing is saved to make framework more reusable.