期刊文献+

由平行平面的投影确定无穷远平面的单应矩阵 被引量:4

Determination of the Infinite Homography Through Scene Parallel Planes
下载PDF
导出
摘要 在三维计算机视觉中,无穷远平面的单应矩阵扮演了极其重要的角色,可使众多视觉问题的求解得到简化.主要讨论如何利用平行平面的投影来求解两个视点间的无穷远平面的单应矩阵,用代数方法构造性地证明了下述结论:(1) 如果场景中含有一组平行平面,则可以通过求解一个一元4次方程来确定两个视点间的无穷远平面对应的单应矩阵;(2) 如果场景中含有两组平行平面,则可以线性地确定两个视点间的无穷远平面对应的单应矩阵.并对上述结果给出了相应的几何解释和具体算法.所给出的结果在三维计算机视觉,特别是摄像机自标定中具有一定的理论意义和应用价值. The homography induced by the plane at infinity between two images, namely the infinite homography, plays a very important role in 3D computer vision since many vision problems could be substantially simplified by knowing it. Unlike homographies induced by ordinary planes which can usually be determined by correspondences of image points, the infinite homography must be determined indirectly since no real physical points lie on the plane at infinity. In this paper, how to determine the infinite homography through scene parallel planes is studied, and the following two conclusions are proved: (1) If only a pair of parallel planes is present in the scene, the infinite homography can be obtained by solving a 4th order polynomial, and at maximum, four possible solutions exist. (2) If at least two pairs of parallel planes exist in the scene, and if planes in different pairs are not parallel, then the infinite homograpgy can be linearly and uniquely determined. In addition, a geometric interpretation to the above results, and some practical algorithms are also provided. The proposed results in the paper are of interests in camera self-calibration and image based 3D reconstruction under both theoretical and practical standpoints.
出处 《软件学报》 EI CSCD 北大核心 2003年第5期936-946,共11页 Journal of Software
基金 国家自然科学基金 中国科学院重大交叉学科前沿资助项目~~
关键词 三维计算机视觉 图像匹配 平行平面 投影 无穷远平面 单应矩阵 infinite homography fundamental matrix camera抯 intrinsic parameter
  • 相关文献

参考文献1

二级参考文献11

  • 1Du F L,Proc IEEE Conference Computer Vision Pattern Recognition,1993年,477页
  • 2吴福朝,自动化学报,2001年,27卷,6期,736页
  • 3Avidan S,Proc European Conference on Computer Vision,1998年,124页
  • 4Quan L,IEEE Trans Pattern Analysis Machine Intelligence,1997年,19卷,8期,834页
  • 5Ma S D,IEEE Trans Robotics Automation,1996年,12卷,1期,114页
  • 6Quan L,Int J Computer Vision,1996年,19卷,1期,93页
  • 7Li F X,Proc European Conference on Computer Vision LNCS 1064/5,1996年,156页
  • 8Shashua A,IEEE Trans Pattern Analysis Machine Intelligence,1994年,16卷,8期,778页
  • 9Li M X,Proc European Conference on Computer Vision LNCS 800/801,1994年,543页
  • 10杨长江,汪威,胡占义.一种基于主动视觉的摄像机内参数自定标方法[J].计算机学报,1998,21(5):428-435. 被引量:38

共引文献32

同被引文献39

  • 1周贤波,冯龙龄.基于DSP和FPGA图像采集技术的研究[J].光学技术,2006,32(z1):141-143. 被引量:5
  • 2黄凤荣,胡占义,吴毅红.单幅图像测量的一种新方法(英文)[J].自动化学报,2004,30(4):487-495. 被引量:10
  • 3章毓晋.图象处理和分析[M].北京:清华大学出版社,1999..
  • 4Faugeras O.Stratification of 3-D vision:projective,affine and metric representations[J].Journal of Optical Society of America,1995,12(3):465-484
  • 5Saspi Y,Irani M.Spatio-temporal alignment of sequences[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(11):1409-1424
  • 6Guo F,Chellapa R.Video mensuration using a stationary camera[C]//Proceedings of the 9th European Conference on Computer Vision,Graz,2006:164-176
  • 7Chen J,Dawson M D,Dixon E W,et al.Adaptive homography-based visual servo tracking for a fixed camera configuration with a camera-in-hand extension[J].IEEE Transactions on Control System Technology,2005,13(5):814-825
  • 8Fang Y,Dixon E W,Dawson M D,et al.Homography-based visual servo regulation of mobile robots[J].IEEE Transactions on System,Man and Cybernetics Part B:Cybernetics,2005,35(5):1041-1050
  • 9Hartley R,Zisserman A.Multiple view geometry in computer vision[M].London:Cambridge University Press,2000:316
  • 10Zhang Z.Flexible camera calibration by viewing a plane from unknown orientation[C]//Proceedings of the 7th International Conference on Computer Vision,Corfu,1999:666-673

引证文献4

二级引证文献18

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部