Mipmap映射技术中d值计算方法的探讨

在Mipmap技术中，常常用一个正方形来近似表示象素在纹理空间的投影，通过正方形的边长d与Mipmp标准采样使用的正方形过滤器的D进行比较，查找Mipmap表。由于象素的投影常是不规则的，求解这个近似正方形的边长d也是不容易的。文中分析了Mipmap预过滤的原理和常用d值的求取办法，提出了一个采用四边形对角线求取近似正方形边长d的方法，并对该方法进行了探讨。

Study on the method of computing d in Mipmap mapping technique

In Mipmap technique, the projection in texture space of a pixel is often approximately expressed as a square. Mipmap table is looked up by comparing side length d of the square with D of the square filter which is used in standard sampling of Mipmap. Computing the side length d of the approximate square is not easy as the projection of the pixel often is irregular. The principle of Mipmap pre-filtration was analyzed and ways of evaluation in common use of d, and a method for computing the side length d of approximate square via diagonal of quadrangle were presented. The discussion on the efficiency of the method was also included.