一种快速生成预定地形高程数据的方法

Fast Algorithm for Scheduled Terrain Generation

针对分形、Perlin噪声不能精确生成预定地形高程数据的问题,先使用分形和高度面结合的生成方法在一定程度上控制生成地形的形状,接着利用一种基于二维高斯分布曲面的变形方法对地形进行变形,然后建立变形的目标函数和搜索空间,用四叉分割树来减少搜索空间,并利用模拟退火法搜索出满意解来实现对预定地形高程的生成。最后提出一种基于顶点影响力分层的算法改进搜索效率,仿真实验表明,该方法快速有效的生成了预定地形。

The method based on fractal and Perlin could not elevation control was applied to generate scheduled terrain generate scheduled terrain, and a method based on fractal and in a large scope, A terrain deformation method based on 2D Gaussian distribution was chosen to deform terrain. The goal function and search space were established, with Quad tree to reduce search space. In order to generate scheduled terrain map, stochastic local search （SLS） was applied to search for optimal solution. Then an algorithm based on the vertex influence was proposed to improve the efficiency of this search. With this method, a scheduled terrain was generated fast and efficiently by simulation.