【24h】

3D modeling tool using a pair of images taken from arbitrary viewpoints

机译:使用从任意视点拍摄的一对图像的3D建模工具

获取原文
获取原文并翻译 | 示例
           

摘要

Several 3D modeling tools using multiple Images taken from arbitrary Viewpoints are proposed. However, there is a great gap between these tools and full automation because most of those are required to determine manually corresponding points. Therefore, this paper proposes a efficient approach in order to restrict the manual procedures to a minimum on 3D modeling. The proposed approach allows to reconstruct 3D structures accurately by discriminating texture regions suitable for correlation matching and uniform color regions suitable for the feature point matching on region contours.
机译:提出了几种使用从任意视点获取的多个图像的3D建模工具。但是,这些工具与完全自动化之间存在很大的差距,因为其中大多数工具需要手动确定相应的点。因此,本文提出了一种有效的方法,以便将手动过程限制在3D建模的最低限度。所提出的方法允许通过区分适合于相关性匹配的纹理区域和适合于区域轮廓上的特征点匹配的均匀颜色区域来精确地重建3D结构。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号