摘  要

地形绘制是三维地形场景绘制的重要组成部分，在计算机动画、游戏动漫等许多领域具有重要的应用价值。本文针对三维地形场景智能生成问题展开研究。首先，介绍了地形场景生成的背景和意义，阐述了当前地形场景生成方法存在的问题和挑战。然后，提出了一种基于深度学习和遗传算法的智能生成方法。该方法通过利用深度学习技术对现有地形数据进行学习和分析，提取地形特征，并通过遗传算法进行优化和调整，生成符合要求的地形场景。最后，通过实验验证了该方法的有效性和可行性。实验结果表明，该方法能够生成高质量、多样化的地形场景，具有较好的实用价值和应用前景。

本文重点研究基于高程图技术和OpenGL技术的三维地形绘制方法。首先，运用LOD技术生成高程数据， 并做平滑处理。 再对高程数据着色后利用OpenGL纹理映射技术实现三维地形的绘制输出。上述方法的实验结果取得预期效果， 基本符合工程需要。

关键词：地形绘制；高程图；LOD；计算机仿真；三维地形

ABSTRACT

Topographic drawing is an important part of 3 D terrain scene drawing, and has important application value in many fields such as computer animation and game animation. This paper studies the problem of intelligent generation of 3 D terrain scenes. First, the background and significance of terrain scene generation are introduced, and the problems and challenges of the current terrain scene generation method are expounded. Then, an intelligent generation method based on deep learning and genetic algorithms. This method learns and analyzes existing topographic data using deep learning techniques, extracts topographic features, and optimizes and adjusts through genetic algorithm to generate topographic scenarios that meet the requirements. Finally, the validity and feasibility of this method are verified by experiments. The experimental results show that this method can generate high quality and diverse topographic scenarios, which has good practical value and application prospect.

This paper focuses on the 3 D topographic mapping method based on the elevation mapping technique and the OpenGL technology. First, the LOD technology is used to generate the elevation data and smooth it out. After coloring the elevation data, the OpenGL texture mapping technology is used to map the output of the 3 D terrain. The experimental results of the above methods achieve the expected results and basically meet the needs of engineering.

Key words: terrain map; elevation map; LOD; computer simulation; 3-dimensional terrain