Paged Cache Based Massive Terrain Dataset Real-Time Rendering Algorithm

This paper presents real-time rendering algorithm of massive terrain dataset based on paged cache. This paper improves the least recently used (LRU) replacement algorithm to swap terrain cache pages, and uses a kind of quadtree terrain pyramid model to organize and manage terrain massive dataset. And real-time loading of terrain data is implemented using view frustum culling and rapid research algorithm for target page based on the resolution test. At the same time, in order to further improve efficiency and parallelism,we use multithreaded programming to read data and generate terrain patch. The experimental results show that usage of cache technology can improve real-time interaction for large scale scene rendering, and the application of multiple threads lets it more fluent.