基于Mipmap的大规模地形绘制算法与仿真  

A Rendering Algorithm and Simulation for Large Scale Terrain based on Mipmap

作　　者：时钢[1] 

机构地区：[1]徐州工业职业技术学院信息工程系,江苏徐州221140

出　　处：《计算机仿真》2015年第2期270-274,共5页Computer Simulation

摘　　要：研究大规模地形与纹理数据动态可视化算法与仿真问题。针对大规模地形与纹理数据动态可视化实时显示难题,提出一种基于Mipmap的大规模地形动态绘制算法,实现模型数据库组织与管理的高效性、场景调度的实时性,以北京市地形数据为仿真模型,实现大规模地形与纹理数据动态可视化算法与实时仿真。上述算法在Mipmap技术的基础上,首先利用双线性内插算法把离散的原始数据生成规则网格DEM,然后采用四叉树结构的LOD算法简化海量地形和纹理数据,最后使用四队列分页调度、基于视域的LRU算法释放内存、多线程并行处理的场景优化策略来进行动态调度和绘制。结合OSG和Open GL渲染引擎,将基于Mipmap的大规模地形绘制算法应用在北京市地形绘制中,仿真结果表明,上述算法比传统的CLOD地形瓦片绘制算法,具有更快的渲染绘制速率,并且能够快速绘制北京市地形数据和在不同场景下进行快速浏览。This paper studies large - scale dynamic terrain and texture data visualization algorithms and simulation problems. Because it is hard to display large -scale terrain and dynamic visual texture data in real time, this paper proposes a rendering algorithm for large scale terrain based on mipmap, which achieves efficient organization and management of model database, and real - time scheduling scenarios. And large - scale dynamic terrain and texture data visualization algorithms and real - time simulation are achieved with the terrain data of Beijing City. Firstly the discrete bilinear interpolation algorithm is used to generate regular grid DEM, then LOD algorithm based on quadtree structure is used to simplify massive terrain and texture data, and finally four queues paging , freeing memory LRU algorithm based on sight, multi -threaded parallel processing scene optimization strategies are used to do dynamic scheduling and mapping. This paper implements the algorithm with Beijing City data, combining with OSG and OpenGL rendering engine. The simulation results show that the algorithm drawn with faster rendering speed than tra- ditional CLOD terrain tiles rendering algorithm, and can make draw Beijing terrain data quickly and switch in differ- ent scenarios quickly.

关 键 词：大规模地形 纹理映射 动态可视化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]