基于网格显著度的视点质量计算方法*

从模糊贴近度的角度，结合网格显著度概念，给出了一种新的视点质量计算方法。首先将场景中的每个三角形看做论域中的一个样本，从可见度大小的角度定义隶属函数，为视点构造一个模糊集合；然后结合三角形的几何面积及面片显著度给出一个对应理想视点的模糊集；最后求取两个模糊集合的贴近度作为视点质量的评价准则。实验结果表明，该算法能够较好地反映视点质量。

Viewpoint quality computing method based on mesh saliency

This paper put forward a new method , which combined the notion of mesh saliency and fuzzy nearness. First, it constructed a fuzzy set for the viewpoint. The elements of the set were faces in the scene, and based membership function on the visibility of faces. Then, defined a ideal fuzzy set, the membership degree was in proportion to the mesh saliency weighted fact area of the face. Finally, took the fuzzy nearness of these two fuzzy set as the measure of viewpoint quality. The experiment results show that this method works well.