Outlier Removal by Convex Optimization for L-Infinity Approaches

  • Yongduek Seo
  • Hyunjung Lee
  • Sang Wook Lee
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5414)

Abstract

This paper is about removing outliers without iterations in L  ∞  optimization. Existing L  ∞  outlier removal method requires iterative removal of the set of measurements with greatest residual during L  ∞  minimization. In the method presented in this paper, on the other hand, a threshold is preset once for the maximum residual error in a manner similar to RANSAC, and the measurements yielding greater residuals than the threshold are taken to be outliers. We examine two feasibility test algorithms: 1) one that minimizes the maximum infeasibility and 2) the other that minimizes the sum of infeasibilities (SOI). Both of these can be used for feasibility test in conjunction with the bisection algorithm which attains the L  ∞  optimum. We note that the SOI method has an interesting characteristic due to its L1-norm minimization nature. It tries to estimate a robust solution while maximizing the number of feasible constraints. The infeasible constraints are found to be due mostly to outliers. Once we set a threshold, the SOI algorithm sorts out outliers from the data set without any repetition and substantial reduction of computation time can be achieved compared to the iterative method. Experiments with synthetic as well as real objects demonstrate the effectiveness of the SOI method. We suggest that the SOI method precede the outlier-sensitive L  ∞  optimization.

References

  1. 1.
    Hartley, R., Schaffalitzky, F.: L  ∞  minimization in geometric reconstruction problems. In: Proc. IEEE Conf. Computer Vision and Pattern Recognition (2004)Google Scholar
  2. 2.
    Kahl, F.: Multiple view geometry and the L  ∞ -norm. In: Proc. Int. Conf. on Computer Vision, Beijing, China, pp. 1002–1009 (2005)Google Scholar
  3. 3.
    Ke, Q., Kanade, T.: Quasiconvex optimization for robust geometric reconstruction. In: Proc. Int. Conf. on Computer Vision, Beijing, China (2005)Google Scholar
  4. 4.
    Sim, K., Hartley, R.: Recovering camera motion using L  ∞  minimization. In: Proc. IEEE Conf. Computer Vision and Pattern Recognition (2006)Google Scholar
  5. 5.
    Aström, K., Enquist, O., Olsson, C., Kahl, F., Hartley, R.: An L-infinity approach to structure and motion problems for 1d-vision. In: IEEE International Conference on Computer Vision (2007)Google Scholar
  6. 6.
    Salzmann, M., Hartley, R., Fua, P.: Convex optimization for deformable surface 3-d tracking. In: IEEE International Conference on Computer Vision (2007)Google Scholar
  7. 7.
    Sim, K., Hartley, R.: Removing outliers using the L  ∞  norm. In: Proc. IEEE Conf. Computer Vision and Pattern Recognition (2006)Google Scholar
  8. 8.
    Li, H.: A practical algorithm for L  ∞  triangulation with outliers. In: IEEE Internatonal Conference on Computer Vision and Pattern Recognition (2007)Google Scholar
  9. 9.
    Olsson, C., Eriksson, A.P., Kahl, F.: Efficient optimization for L  ∞  problems using pseudoconvexity. In: IEEE International Conference on Computer Vision (2007)Google Scholar
  10. 10.
    Seo, Y., Hartley, R.I.: A fast method to minimize L  ∞  error norm for geometric vision problems. In: IEEE International Conference on Computer Vision (2007)Google Scholar
  11. 11.
    Fischler, M.A., Bolles, R.C.: Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography. Comm. of the ACM, 381–395 (1981)Google Scholar
  12. 12.
    Sturm, J.: Using SeDuMi 1.02, a Matlab toolbox for optimization over symmetric cones. Optimization Methods and Software 11-12, 625–653 (1999)MathSciNetCrossRefMATHGoogle Scholar
  13. 13.
    Hartley, R., Kahl, F.: Global optimization through searching rotation space and optimal estimation of the essential matrix. In: IEEE International Conference on Computer Vision (2007)Google Scholar
  14. 14.
    Boyd, S., Vandenberghe, L.: Convex Optimization. Cambridge University Press, Cambridge (2004)CrossRefMATHGoogle Scholar
  15. 15.
    Press, W., Teukolsky, S., Vetterling, W., Flannery, B.: Numerical Recipes in C: The Art of Scientiic Computing. Cambridge University Press, Cambridge (1992)MATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Yongduek Seo
    • 1
  • Hyunjung Lee
    • 1
  • Sang Wook Lee
    • 1
  1. 1.Department of Media TechnologySogang UniversityKorea

Personalised recommendations