摘要
现有的递进网格生成算法不仅效率低,而且大多只能完成几何特征的简化,没有考虑网格模型的其它表面属性.针对这些问题,提出一种新算法.该算法把基于点到一组平面距离平方和的二次误差测度,从三维几何空间推广到包含属性信息的多维空间.这种二次误差测度表示了简化网格与初始网格的几何特征和属性信息的匹配程度,利用二次误差值最小原理,指导网格简化操作的进行.实验结果表明,该算法不仅效率高,而且可以保证简化模型同初始模型在几何特征和颜色信息上具有高相似度.
Current algorithms for generating progressive meshes often have a low efficiency and can only complete the simplification of the geometric characters, without considering the other surface attributes. In order to solve these problems, we present a new algorithm in this paper. This algorithm generalizes the quadrie error metric, which is based on the sum of squared distances of the vertex to all the planes in its set, from 3 - dimension to n - dimension including attributes information. Such a quadric error metric denotes the matching degree of the geometric characters and attributes information between the simplified mesh and the original one and then directs the process of mesh sim- plification under the least quadric error principle. The experimental results prove that this algorithm not only has a high efficiency, but also can guarantee a high resemblance of the geometric characters and color information between the simplified model and the original model.
出处
《郑州大学学报（工学版）》
CAS
2006年第2期84-87,共4页
Journal of Zhengzhou University（Engineering Science）
基金
河南省科技攻关项目(0324210045)
关键词
二次误差测度
递进网格
边折叠
细节层次
quadric error metric
progressive mesh
edge collapse
level of detail
