• DocumentCode
    2480616
  • Title

    Critical configurations of lines to geometry determination of three cameras

  • Author

    Zhao, Ming ; Chung, Ronald

  • Author_Institution
    Dept. of Mech. & Autom. Eng., Chinese Univ. of Hong Kong, Hongkong
  • fYear
    2008
  • fDate
    8-11 Dec. 2008
  • Firstpage
    1
  • Lastpage
    5
  • Abstract
    We address the question of, what structure of a set of lines in space constitutes a critical configuration to three generally positioned cameras. By critical configuration, it is meant a set of lines in space whose image projections to the cameras do not allow unique determination of the cameraspsila relative positions in space. We approach the question by looking into the trifocal tensor of the cameraspsila a quantity tightly related to the camera geometry. We focus on structures of lines in space that are common in reality - the linear structures - and examine which of them would not allow the tensor to be uniquely determined. Linear structures of lines include linear ruled surface, linear line congruence, and linear line complex, and more specifically line pencil, point star, and ruled plane. We offer a summary of by how much such families of line set are shy of the general enough structure for a unique determination of the tensor. We also point out how many lines need be visible minimally in each of them, for the full information of the associated structure to be revealed in the image data. We present synthetic and real image results to confirm the findings. The findings are important to the validity and stability of algorithms related to structure from motion and projective reconstruction using line correspondences.
  • Keywords
    cameras; computational geometry; image classification; surface fitting; tensors; camera geometry; critical configuration; image projection; line geometry determination; linear line complex; linear line congruence; linear line structure; linear ruled surface; rank classification; trifocal tensor; Automation; Cameras; Computational geometry; Content addressable storage; Image reconstruction; Orbital robotics; Robot vision systems; Space technology; Stability; Tensile stress;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Pattern Recognition, 2008. ICPR 2008. 19th International Conference on
  • Conference_Location
    Tampa, FL
  • ISSN
    1051-4651
  • Print_ISBN
    978-1-4244-2174-9
  • Electronic_ISBN
    1051-4651
  • Type

    conf

  • DOI
    10.1109/ICPR.2008.4761368
  • Filename
    4761368