复杂几何模型的混合绘制算法研究

An Approach of Image Modeling for Large Scale Geometric Models

基于几何与图像的混合建模与绘制是解决复杂几何模型快速绘制的一个有效途径 .提出一种复杂几何模型的混合绘制方法 ,首先给出基于几何投影的 Warp变换公式 ,然后对几何模型进行预绘制 ,得到具有深度的图像 ;依据图像分辨率对该图像进行几何重构 ,得到具有几何拓扑关系的图像模型 ,并对该模型进行进一步修正 ;最后解决了模型的冗余检测问题 .实验结果表明 。

How to realize fast rendering of large scale geometric models has always been a difficult problem in computer graphics. Geometry and image based hybrid modeling and rendering is an effective solution to this problem. This paper formulates four equations of 3D warp transformation under geometric projection. Geometric models are pre rendered to get images with recorded depth. Geometry reconstruction is performed from images taking into consideration the distance among pixels to keep the correct topological relationship of original images. The resultant models are refined to eliminate the overlapping redundancy of different objects. Experiment results illustrate that this hybrid modeling approach can realize fast rendering of geometry without losing the accuracy.