Advertisement

The Distinction between Virtual and Physical Planes Using Homography

  • A. Amintabar
  • Boubaker Boufama
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5627)

Abstract

Identification of planes from a pair of uncalibrated stereo images is a challenging problem as it can lead to extracting virtual planes instead of physical ones, especially for complex scenes. We propose a new homography-based approach to extract physical planes and to distinguish them from virtual ones for general scenarios. The proposed approach uses noncoplanar points inside a plane to decide whether the plane is physical or virtual. Depending on the distribution of the points inside the convex hull of the plane, the plane is classified as likely virtual, likely physical, very likely virtual or very likely physical. To estimate homographies, we use our method which computes the homography for three points with no necessity to assume the fourth point being coplanar with the three. Experiments on real images demonstrate the validity of the proposed approach for general scenarios.

Keywords

Stereo images homography feature points virtual plane 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Wang, G., Wu, Q.M.J., Zhang, W.: Kruppa equation based camera calibration from homography. Pattern Recognition Letters archive 29(16), 2137–2144 (2008)CrossRefGoogle Scholar
  2. 2.
    Simond, N., Rives, P.: What can be done with an embedded stereo-rig in urban environments. Robotics and Autonomous Systems 56(9), 777–789 (2008)CrossRefGoogle Scholar
  3. 3.
    Kirchhof, M.: Linear Constraints in Two-View Multiple Homography Estimation of Uncalibrated Scenes. In: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. ISPRS 2008, pp. 13–20 (2008)Google Scholar
  4. 4.
    Wang, G., Tsui, T., Hu, Z.: Reconstruction of structured scenes from two uncalibrated images. Pattern Recognition Letters 26(2), 207–220 (2005)CrossRefGoogle Scholar
  5. 5.
    He, Q., Chu, C.H.: Planar Surface Detection in Image Pairs Using Homographic Constraints. In: Bebis, G., Boyle, R., Parvin, B., Koracin, D., Remagnino, P., Nefian, A., Meenakshisundaram, G., Pascucci, V., Zara, J., Molineros, J., Theisel, H., Malzbender, T. (eds.) ISVC 2006. LNCS, vol. 4291, pp. 19–27. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  6. 6.
    Kawakam, H., Ito, Y., Kanazawa, Y.: A robust method for detecting planar regions based on random sampling. Systems and Computers in Japan 37(4), 11–22 (2006)CrossRefGoogle Scholar
  7. 7.
    Zeng, H., Deng, X., Hu, Z.: A new normalized method on line-based homography estimation. Pattern Recognition Letters 29(9), 1236–1244 (2008)CrossRefGoogle Scholar
  8. 8.
    Choi, O., Kim, H., Kweon, I.S.: Simultaneous Plane Extraction and 2D Homography Estimation Using Local Feature Transformations. In: Yagi, Y., Kang, S.B., Kweon, I.S., Zha, H. (eds.) ACCV 2007, Part II. LNCS, vol. 4844, pp. 269–278. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  9. 9.
    Hartley, R., Zisserman, A.: Multiple View Geometry in Computer Vision. Cambridge University Press, Cambridge (2003)zbMATHGoogle Scholar
  10. 10.
    Zhang, Z., Deriche, R., Faugeras, O., Luong, Q.T.: A robust technique for matching two uncalibrated images through the recovery if the unknown epipolar geometry. Artificial Intelligence 78, 87–119 (1995)CrossRefGoogle Scholar
  11. 11.
    Chan, T.M.: Optimal output-sensitive convex hull algorithms in two and three dimensions. Discrete and Computational Geometry 16, 361–368 (1996)MathSciNetCrossRefzbMATHGoogle Scholar
  12. 12.
    Amintabar, A., Boufama, B.: Homography-based plane identification and matching. In: ICIP 2008. In: 15th IEEE International conference on Image Processing, San Diego, California, U.S.A, October 12–15, 2008, pp. 297–300 (2008)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • A. Amintabar
    • 1
  • Boubaker Boufama
    • 1
  1. 1.School of Computer ScienceUniversity of WindsorWindsorCanada

Personalised recommendations