计算机与现代化

• 算法设计与分析 • 上一篇    下一篇

基于特征匹配的关联投影纹理映射

  

  1. 中国人民解放军理工大学指挥信息系统学院,江苏南京210007
  • 收稿日期:2015-10-22 出版日期:2016-01-22 发布日期:2016-01-26
  • 作者简介:刘子才(1990-),男,广东云浮人,中国人民解放军理工大学指挥信息系统学院硕士研究生,研究方向:计算机网络,图形学; 胡谷雨(1963-),男,浙江东阳人,教授,博士生导师,研究方向:计算机网络,智能网络; 罗建欣(1984-),男,江苏南京人,讲师,博士,研究方向:计算机网络,栅格化建模。
  • 基金资助:
    国家863计划项目(2012AA01A509); 江苏省自然科学青年基金资助项目(BK20150722)

Relation Projection Texture Mapping Based on Feature Matching

  1. Academy of Command Information System, PLA University of Science and Technology, Nanjing 210007, China
  • Received:2015-10-22 Online:2016-01-22 Published:2016-01-26

摘要: 在虚拟现实场景中,动态现实的重建过程尤为复杂,一般需要大量包含现实信息的图像进行整合。这些图像大部分都与其中一张或多张图像存在着公共区域,本文提出基于特征匹配的关联投影映射方法,根据这些公共区域,通过SIFT特征匹配提取图像之间的关联特性,利用此关联构造视图变换矩阵进行新的投影纹理映射。实验结果表明,这种方法稳定有效,并在重建动态现实中具有良好的应用效果。

关键词: 现实, SIFT特征, 关联, 投影纹理映射

Abstract: In virtual reality scenes, the reconstruction of the dynamic reality process is complicated that needs a lot of images which contain real information to consolidate. These images most have a public area with one or more images. This paper puts forward an algorithm named relation projection mapping method based on feature matching. On the basis of these public areas, it can extract the associated features among the images by SIFT feature matching. With the associated features, we can proceed a new projective texture mapping by constructing the view transformation matrix. The experimental results show that this method is stable and effective, and the dynamic reconstruction has a good application effect in reality.

Key words: reality, SIFT feature, relevance, projective texture mapping

中图分类号: