基于投影网格的大面积水面生成算法

Real-time water rendering using projected grid

本文主要研究在现代硬件条件下,如何为渲染生成一个大面积的水面。为了达到摄像机高低不同时对分辨率的要求,同时保持空间的可量测性,现在经常使用的生成算法是各种"层次细节"的方法。但层次细节也存在存储数据量大以及接边等问题。本文介绍了一种替代的算法:投影网格算法。投影网格从我们希望在投影后得到的网格入手,反向找出网格在投影前的位置,更有目的性,有自己的优势。

This thesis will examine how a large water surface can be rendered in an efficient manner using modem graphics hardware. We often treat the water surface as a height field for simplification. To allow spatial scalabillty, different methods of ＂Level-Of- Detail＂ （LOD） are often used. The thesis presents an ahemative technique called ＂projected grid＂ . It creates a grid mesh whosevertices are even-spaced, not in world-space, which is the traditional way but in post-perspective camera space. This will deliver a polygonal representation that provides spatial scalability along with high relative resolution without resorting to multiple levels of detail.