首页 | 官方网站   微博 | 高级检索  
     

基于轮廓线的三维重建方法的研究
引用本文:吴涛,刘金义.基于轮廓线的三维重建方法的研究[J].辽宁石油化工大学学报,1999,19(3):44-48.
作者姓名:吴涛  刘金义
作者单位:抚顺石油学院计算机科学与技术系,辽宁抚顺,113001
摘    要:在几何造型系统中引入基于轮廓线的三维重建方法作为造型手段的补充是有意义的。实际应用中得到的轮廓线数据集所具有的不同特点,使得三维重建过程中要处理的问题域被分为轮廓线的二维投影域和三维轮廓线整体信息域两种情况。对于轮廓的二维投影域 可以使用 Delaunay 三 角剖分二维任意域的算法来处理,并且在应用中对于原算法存在的缺陷进行了修正,提高了算法的健壮性;处理三维轮廓线整体信息域时,可以考虑使用基于图论描述的组合优化 的求解方法加以解决,在多个 三维重建结果中选择与定义最接近的解,是比较合理的解决问题的思路,方法中涉及优化目标和准则的选择,优化算法的运用以及三维重建结果向现有的几何造型系统使用的数据结构的转化算法等问题

关 键 词:三维重建    轮廓线    几何造型    计算机图形学
修稿时间:1998-09-07

Study on Three-Dimensional Reconstruction Based on Contour Data
Wu Tao,Liu Jinyi.Study on Three-Dimensional Reconstruction Based on Contour Data[J].Journal of Liaoning University of Petroleum & Chemical Technology,1999,19(3):44-48.
Authors:Wu Tao  Liu Jinyi
Abstract:It is useful to introduce the three-dimensional reconstruction as a supplement of modeling method into geometry model system. As to the characteristic of different contour data, the problem of reconstruction is classified into two categories, one is the reconstruction base on two-dimensional projection of contour data, and the other is the reconstruction based on the three-dimensional contour data. The former situation can be solved by Delaunay Triangulation algorithm on two-dimensional arbitrary domain, and the robust of algorithm is improved by correcting the bug in algorithm. The solution of the latter situation is the method to combinatorial optimization by use of diagram theory. It is reasonable that selecting a result which is closed to the definition from many results. During the method, the attention is focused on the selection of optimization goals and rules, application of optimization algorithm and the translation algorithm from the result of three-dimensional reconstruction to data structure which is used by geometry model system.
Keywords:Three-dimensional reconstruction  Contour data  Geometry modeling  Computer graphics
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号