三维网格模型的图像化表示  

Graphic Representation of 3D Mesh Model

作　　者：肖雄[1] 杨战军[2] 朱杰[1] 

机构地区：[1]河海大学计算机与信息学院 [2]92995部队

出　　处：《计算机与数字工程》2015年第4期720-724,共5页Computer & Digital Engineering

摘　　要：为使三维网格模型能被GPU进行渲染,提出了一种适用于GPU的针对零亏格的三角网格模型的绘制框架。传统的方法生成几何图像一般是从原始网格逐步切割至平面域,但是这样会产生复杂接缝的问题。论文通过对参数化后的球面进行映射,从而间接生成几何图像,避免了对原始网格进行复杂的切割。首先,将已经球面参数化后的球面信息映射至立方体,立方体平铺开即构成一个二维几何图像;然后,将二维纹理信息传输至GPU,利用GPU来还原三角网格模型。此文采用OpenGL和CUDA相结合的方式来实现最终效果,实验结果表明该绘制框架是可行的,参数化效果和还原效果较好。To make 3D mesh model be rendered by the GPU, a drawing framework of GPU for zero-genus triangular mesh model is proposed. The traditional method of generating geometry image is generally from the original mesh to plane domain by gradually cutting, but it will cause complex joint problems. This article maps based on the sphere which has pa- rameterized, thus generates geometry image indirectly, so as to avoid the complex cutting to the original mesh. First, the tri- angular mesh is spherical parameterized and the spherical information is mapped onto a cube, the cube is laid out to form a 2D geometry image. Then, 2D texture information is transferred to the GPU and the triangular mesh model is restored by GPU. The article combines OpenGL and CUDA to achieve the final effect and the experimental results show that the drawing framework is feasible with good parameterization effect and restoration effect.

关 键 词：三角网格 网格参数化 几何图像 GPU 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]