Abstract

We address the model-to-image registration problem with line features in the following two ways. (a) We present a robust solution to simultaneously recover the camera pose and the three-dimensional-to-two-dimensional line correspondences. With weak pose priors, our approach progressively verifies the pose guesses with a Kalman filter by using a subset of recursively found match hypotheses. Experiments show our method is robust to occlusions and clutter. (b) We propose a new line feature based pose estimation algorithm, which iteratively optimizes the objective function in the object space. Experiments show that the algorithm has strong robustness to noise and outliers and that it can attain very accurate results efficiently.

© 2012 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM 24, 381–395 (1981).
    [CrossRef]
  2. P. David, D. Dementhon, R. Duraiswami, and H. Samet, “SoftPOSIT: Simultaneous pose and correspondence determination,” Int. J. Comput. Vis. 59, 259–284 (2004).
    [CrossRef]
  3. P. David, D. Dementhon, R. Duraiswami, and H. Samet, “Simultaneous pose and correspondence determination using line features,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003 (IEEE, 2003), pp. II-424–II-431.
  4. F. Moreno-Noguer, V. Lepetit, and P. Fua, “Pose priors for simultaneously solving alignment and correspondence,” in Proceedings of the 10th European Conference on Computer Vision: Part II (Springer-Verlag, 2008), pp. 405–418.
  5. M. Dhome, M. Richetin, J. thierry Lapreste, and G. Rives, “Determination of the attitude of 3-D objects from a single perspective view,” IEEE Trans. Pattern Anal. Mach. Intell. 11, 1265–1278 (1989).
    [CrossRef]
  6. D. DeMenthon and L. Davis, “Model-based object pose in 25 lines of code,” Int. J. Comput. Vis. 15, 123–141 (1995).
    [CrossRef]
  7. R. Horaud, F. Dornaika, B. Lamiroy, and S. Christy, “Object pose: the link between weak perspective, paraperspective and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
    [CrossRef]
  8. S. Christy and R. Horaud, “Iterative pose computation from line correspondences,” Comput. Vis. Image Understand. 73, 137–144 (1999).
    [CrossRef]
  9. C.-P. Lu, G. D. Hager, and E. Mjolsness, “Fast and globally convergent pose estimation from video images,” IEEE Trans. Pattern Anal. Mach. Intell. 22, 610–622 (2000).
    [CrossRef]
  10. A. Ansar and K. Daniilidis, “Linear pose estimation from points or lines,” IEEE Trans. Pattern Anal. Mach. Intell. 25, 578–589 (2003).
    [CrossRef]
  11. A. Castro, Y. Frauel, E. Tepichin, and B. Javidi, “Pose estimation from a two-dimensional view by use of composite correlation filters and neural networks,” Appl. Opt. 42, 5882–5890 (2003).
    [CrossRef]
  12. F. Moreno-Noguer, V. Lepetit, and P. Fua, “Accurate non-iterative O(n) solution to the PnP problem,” in Proceedings of the IEEE 11th International Conference on Computer Vision, 2007 (IEEE, 2007), pp. 1–8.
  13. W. E. L. Grimson, “Object recognition by computer: the role of geometric constraints,” Int. J. Comput. Vis. 31, 350–504 (1990).
  14. J. S. Beis and D. G. Lowe, “Indexing without invariants in 3D object recognition,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 1000–1015 (1999).
    [CrossRef]
  15. Y. Lamdan and H. J. Wolfson, “Geometric hashing: a general and efficient model-based recognition scheme,” in Proceedings of the Second International Conference on Computer Vision (IEEE, 1988), pp. 238–249.
  16. P. Wunsch and G. Hirzinger, “Registration of cad models to images by iterative inverse perspective matching,” in Proceedings of the 13th International Conference on Pattern Recognition (IEEE, 1996), pp. 78–83.
  17. J. R. Beveridge and E. M. Riseman, “Optimal geometric model matching under full 3d perspective,” Comput. Vis. Image Understand. 61, 351–364 (1995).
    [CrossRef]
  18. F. Jurie, “Solution of the simultaneous pose and correspondence problem using gaussian error model,” Comput. Vis. Image Understand. 73, 357–373 (1999).
    [CrossRef]
  19. R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
    [CrossRef]
  20. R. Horaud, F. Dornaika, B. Lamiroy, S. Christy, “Object pose: the link between weak perspective, paraperspective, and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
    [CrossRef]
  21. L. Quan and Z. Lan, “Linear N-point camera pose determination,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 774–780 (1999).
    [CrossRef]
  22. H. H. Chen, “Pose determination from line-to-plane correspondences: existence condition and closed-form solutions,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 530–541 (1991).
    [CrossRef]
  23. Y. Liu, T. S. Huang, and O. D. Faugeras, “A linear algorithm for motion estimation using straight line correspondences,” Comput. Vis. Graph. Image Process. 44, 35–57 (1988).
    [CrossRef]
  24. Y. Liu, T. S. Huang, and L. D. Faugeras, “Determination of camera location from 2-D to 3-D line and point correspondences,” IEEE Trans. Pattern Anal. Mach. Intell. 12, 28–37 (1990).
    [CrossRef]
  25. C.-N. Lee and R. M. Haralick, “Statistical estimation for exterior orientation from line-to-line correspondences,” Image Vis. Comput. 14, 379–388 (1996).
    [CrossRef]
  26. R. Kumar and A. R. Hanson, “Robust methods for estimating pose and a sensitivity analysis,” Comput. Vis. Graph. Image Process. 60, 313–342 (1994).
    [CrossRef]
  27. T. Q. Phong, R. Horaud, A. Yassine, and P. D. Tao, “Object pose from 2D to 3D point and line correspondences,” Int. J. Comput. Vis. 15, 225–243 (1995).
    [CrossRef]
  28. N. Navab and O. Faugeras, “Monocular pose determination from lines: critical sets and maximum number of solutions,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (IEEE, 1993), pp. 254–260.
  29. S. Umeyama, “Least-squares estimation of transformatio parameters between two point patterns,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 376–380 (1991).
    [CrossRef]
  30. B. K. P. Horn, H. Hilden, and S. Negahdaripour, “Closed-form solution of absolute orientation using orthonormal matrices,” J. Opt. Soc. Am. 5, 1127–1135 (1988).
    [CrossRef]
  31. H. Wuest, F. Vial, and D. Stricker, “Adaptive line tracking with multiple hypotheses for augmented reality,” in Proceedings of the 4th IEEE/ACM International Symposium on Mixed and Augmented Reality (IEEE, 2005), pp. 62–69.

2004

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “SoftPOSIT: Simultaneous pose and correspondence determination,” Int. J. Comput. Vis. 59, 259–284 (2004).
[CrossRef]

2003

2000

C.-P. Lu, G. D. Hager, and E. Mjolsness, “Fast and globally convergent pose estimation from video images,” IEEE Trans. Pattern Anal. Mach. Intell. 22, 610–622 (2000).
[CrossRef]

1999

L. Quan and Z. Lan, “Linear N-point camera pose determination,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 774–780 (1999).
[CrossRef]

J. S. Beis and D. G. Lowe, “Indexing without invariants in 3D object recognition,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 1000–1015 (1999).
[CrossRef]

F. Jurie, “Solution of the simultaneous pose and correspondence problem using gaussian error model,” Comput. Vis. Image Understand. 73, 357–373 (1999).
[CrossRef]

S. Christy and R. Horaud, “Iterative pose computation from line correspondences,” Comput. Vis. Image Understand. 73, 137–144 (1999).
[CrossRef]

1997

R. Horaud, F. Dornaika, B. Lamiroy, and S. Christy, “Object pose: the link between weak perspective, paraperspective and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, S. Christy, “Object pose: the link between weak perspective, paraperspective, and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

1996

C.-N. Lee and R. M. Haralick, “Statistical estimation for exterior orientation from line-to-line correspondences,” Image Vis. Comput. 14, 379–388 (1996).
[CrossRef]

1995

T. Q. Phong, R. Horaud, A. Yassine, and P. D. Tao, “Object pose from 2D to 3D point and line correspondences,” Int. J. Comput. Vis. 15, 225–243 (1995).
[CrossRef]

J. R. Beveridge and E. M. Riseman, “Optimal geometric model matching under full 3d perspective,” Comput. Vis. Image Understand. 61, 351–364 (1995).
[CrossRef]

D. DeMenthon and L. Davis, “Model-based object pose in 25 lines of code,” Int. J. Comput. Vis. 15, 123–141 (1995).
[CrossRef]

1994

R. Kumar and A. R. Hanson, “Robust methods for estimating pose and a sensitivity analysis,” Comput. Vis. Graph. Image Process. 60, 313–342 (1994).
[CrossRef]

1991

S. Umeyama, “Least-squares estimation of transformatio parameters between two point patterns,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 376–380 (1991).
[CrossRef]

H. H. Chen, “Pose determination from line-to-plane correspondences: existence condition and closed-form solutions,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 530–541 (1991).
[CrossRef]

1990

W. E. L. Grimson, “Object recognition by computer: the role of geometric constraints,” Int. J. Comput. Vis. 31, 350–504 (1990).

Y. Liu, T. S. Huang, and L. D. Faugeras, “Determination of camera location from 2-D to 3-D line and point correspondences,” IEEE Trans. Pattern Anal. Mach. Intell. 12, 28–37 (1990).
[CrossRef]

1989

M. Dhome, M. Richetin, J. thierry Lapreste, and G. Rives, “Determination of the attitude of 3-D objects from a single perspective view,” IEEE Trans. Pattern Anal. Mach. Intell. 11, 1265–1278 (1989).
[CrossRef]

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

1988

Y. Liu, T. S. Huang, and O. D. Faugeras, “A linear algorithm for motion estimation using straight line correspondences,” Comput. Vis. Graph. Image Process. 44, 35–57 (1988).
[CrossRef]

B. K. P. Horn, H. Hilden, and S. Negahdaripour, “Closed-form solution of absolute orientation using orthonormal matrices,” J. Opt. Soc. Am. 5, 1127–1135 (1988).
[CrossRef]

1981

M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM 24, 381–395 (1981).
[CrossRef]

Ansar, A.

A. Ansar and K. Daniilidis, “Linear pose estimation from points or lines,” IEEE Trans. Pattern Anal. Mach. Intell. 25, 578–589 (2003).
[CrossRef]

Beis, J. S.

J. S. Beis and D. G. Lowe, “Indexing without invariants in 3D object recognition,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 1000–1015 (1999).
[CrossRef]

Beveridge, J. R.

J. R. Beveridge and E. M. Riseman, “Optimal geometric model matching under full 3d perspective,” Comput. Vis. Image Understand. 61, 351–364 (1995).
[CrossRef]

Bolles, R. C.

M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM 24, 381–395 (1981).
[CrossRef]

Castro, A.

Chen, H. H.

H. H. Chen, “Pose determination from line-to-plane correspondences: existence condition and closed-form solutions,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 530–541 (1991).
[CrossRef]

Christy, S.

S. Christy and R. Horaud, “Iterative pose computation from line correspondences,” Comput. Vis. Image Understand. 73, 137–144 (1999).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, S. Christy, “Object pose: the link between weak perspective, paraperspective, and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, and S. Christy, “Object pose: the link between weak perspective, paraperspective and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

Daniilidis, K.

A. Ansar and K. Daniilidis, “Linear pose estimation from points or lines,” IEEE Trans. Pattern Anal. Mach. Intell. 25, 578–589 (2003).
[CrossRef]

David, P.

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “SoftPOSIT: Simultaneous pose and correspondence determination,” Int. J. Comput. Vis. 59, 259–284 (2004).
[CrossRef]

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “Simultaneous pose and correspondence determination using line features,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003 (IEEE, 2003), pp. II-424–II-431.

Davis, L.

D. DeMenthon and L. Davis, “Model-based object pose in 25 lines of code,” Int. J. Comput. Vis. 15, 123–141 (1995).
[CrossRef]

Dementhon, D.

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “SoftPOSIT: Simultaneous pose and correspondence determination,” Int. J. Comput. Vis. 59, 259–284 (2004).
[CrossRef]

D. DeMenthon and L. Davis, “Model-based object pose in 25 lines of code,” Int. J. Comput. Vis. 15, 123–141 (1995).
[CrossRef]

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “Simultaneous pose and correspondence determination using line features,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003 (IEEE, 2003), pp. II-424–II-431.

Dhome, M.

M. Dhome, M. Richetin, J. thierry Lapreste, and G. Rives, “Determination of the attitude of 3-D objects from a single perspective view,” IEEE Trans. Pattern Anal. Mach. Intell. 11, 1265–1278 (1989).
[CrossRef]

Dornaika, F.

R. Horaud, F. Dornaika, B. Lamiroy, S. Christy, “Object pose: the link between weak perspective, paraperspective, and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, and S. Christy, “Object pose: the link between weak perspective, paraperspective and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

Duraiswami, R.

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “SoftPOSIT: Simultaneous pose and correspondence determination,” Int. J. Comput. Vis. 59, 259–284 (2004).
[CrossRef]

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “Simultaneous pose and correspondence determination using line features,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003 (IEEE, 2003), pp. II-424–II-431.

Faugeras, L. D.

Y. Liu, T. S. Huang, and L. D. Faugeras, “Determination of camera location from 2-D to 3-D line and point correspondences,” IEEE Trans. Pattern Anal. Mach. Intell. 12, 28–37 (1990).
[CrossRef]

Faugeras, O.

N. Navab and O. Faugeras, “Monocular pose determination from lines: critical sets and maximum number of solutions,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (IEEE, 1993), pp. 254–260.

Faugeras, O. D.

Y. Liu, T. S. Huang, and O. D. Faugeras, “A linear algorithm for motion estimation using straight line correspondences,” Comput. Vis. Graph. Image Process. 44, 35–57 (1988).
[CrossRef]

Fischler, M. A.

M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM 24, 381–395 (1981).
[CrossRef]

Frauel, Y.

Fua, P.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Pose priors for simultaneously solving alignment and correspondence,” in Proceedings of the 10th European Conference on Computer Vision: Part II (Springer-Verlag, 2008), pp. 405–418.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Accurate non-iterative O(n) solution to the PnP problem,” in Proceedings of the IEEE 11th International Conference on Computer Vision, 2007 (IEEE, 2007), pp. 1–8.

Grimson, W. E. L.

W. E. L. Grimson, “Object recognition by computer: the role of geometric constraints,” Int. J. Comput. Vis. 31, 350–504 (1990).

Hager, G. D.

C.-P. Lu, G. D. Hager, and E. Mjolsness, “Fast and globally convergent pose estimation from video images,” IEEE Trans. Pattern Anal. Mach. Intell. 22, 610–622 (2000).
[CrossRef]

Hanson, A. R.

R. Kumar and A. R. Hanson, “Robust methods for estimating pose and a sensitivity analysis,” Comput. Vis. Graph. Image Process. 60, 313–342 (1994).
[CrossRef]

Haralick, R. M.

C.-N. Lee and R. M. Haralick, “Statistical estimation for exterior orientation from line-to-line correspondences,” Image Vis. Comput. 14, 379–388 (1996).
[CrossRef]

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

Hilden, H.

B. K. P. Horn, H. Hilden, and S. Negahdaripour, “Closed-form solution of absolute orientation using orthonormal matrices,” J. Opt. Soc. Am. 5, 1127–1135 (1988).
[CrossRef]

Hirzinger, G.

P. Wunsch and G. Hirzinger, “Registration of cad models to images by iterative inverse perspective matching,” in Proceedings of the 13th International Conference on Pattern Recognition (IEEE, 1996), pp. 78–83.

Horaud, R.

S. Christy and R. Horaud, “Iterative pose computation from line correspondences,” Comput. Vis. Image Understand. 73, 137–144 (1999).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, S. Christy, “Object pose: the link between weak perspective, paraperspective, and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, and S. Christy, “Object pose: the link between weak perspective, paraperspective and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

T. Q. Phong, R. Horaud, A. Yassine, and P. D. Tao, “Object pose from 2D to 3D point and line correspondences,” Int. J. Comput. Vis. 15, 225–243 (1995).
[CrossRef]

Horn, B. K. P.

B. K. P. Horn, H. Hilden, and S. Negahdaripour, “Closed-form solution of absolute orientation using orthonormal matrices,” J. Opt. Soc. Am. 5, 1127–1135 (1988).
[CrossRef]

Huang, T. S.

Y. Liu, T. S. Huang, and L. D. Faugeras, “Determination of camera location from 2-D to 3-D line and point correspondences,” IEEE Trans. Pattern Anal. Mach. Intell. 12, 28–37 (1990).
[CrossRef]

Y. Liu, T. S. Huang, and O. D. Faugeras, “A linear algorithm for motion estimation using straight line correspondences,” Comput. Vis. Graph. Image Process. 44, 35–57 (1988).
[CrossRef]

Javidi, B.

Joo, H.

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

Jurie, F.

F. Jurie, “Solution of the simultaneous pose and correspondence problem using gaussian error model,” Comput. Vis. Image Understand. 73, 357–373 (1999).
[CrossRef]

Kim, M. B.

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

Kumar, R.

R. Kumar and A. R. Hanson, “Robust methods for estimating pose and a sensitivity analysis,” Comput. Vis. Graph. Image Process. 60, 313–342 (1994).
[CrossRef]

Lamdan, Y.

Y. Lamdan and H. J. Wolfson, “Geometric hashing: a general and efficient model-based recognition scheme,” in Proceedings of the Second International Conference on Computer Vision (IEEE, 1988), pp. 238–249.

Lamiroy, B.

R. Horaud, F. Dornaika, B. Lamiroy, and S. Christy, “Object pose: the link between weak perspective, paraperspective and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, S. Christy, “Object pose: the link between weak perspective, paraperspective, and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

Lan, Z.

L. Quan and Z. Lan, “Linear N-point camera pose determination,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 774–780 (1999).
[CrossRef]

Lee, C.-N.

C.-N. Lee and R. M. Haralick, “Statistical estimation for exterior orientation from line-to-line correspondences,” Image Vis. Comput. 14, 379–388 (1996).
[CrossRef]

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

Lepetit, V.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Accurate non-iterative O(n) solution to the PnP problem,” in Proceedings of the IEEE 11th International Conference on Computer Vision, 2007 (IEEE, 2007), pp. 1–8.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Pose priors for simultaneously solving alignment and correspondence,” in Proceedings of the 10th European Conference on Computer Vision: Part II (Springer-Verlag, 2008), pp. 405–418.

Liu, Y.

Y. Liu, T. S. Huang, and L. D. Faugeras, “Determination of camera location from 2-D to 3-D line and point correspondences,” IEEE Trans. Pattern Anal. Mach. Intell. 12, 28–37 (1990).
[CrossRef]

Y. Liu, T. S. Huang, and O. D. Faugeras, “A linear algorithm for motion estimation using straight line correspondences,” Comput. Vis. Graph. Image Process. 44, 35–57 (1988).
[CrossRef]

Lowe, D. G.

J. S. Beis and D. G. Lowe, “Indexing without invariants in 3D object recognition,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 1000–1015 (1999).
[CrossRef]

Lu, C.-P.

C.-P. Lu, G. D. Hager, and E. Mjolsness, “Fast and globally convergent pose estimation from video images,” IEEE Trans. Pattern Anal. Mach. Intell. 22, 610–622 (2000).
[CrossRef]

Mjolsness, E.

C.-P. Lu, G. D. Hager, and E. Mjolsness, “Fast and globally convergent pose estimation from video images,” IEEE Trans. Pattern Anal. Mach. Intell. 22, 610–622 (2000).
[CrossRef]

Moreno-Noguer, F.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Accurate non-iterative O(n) solution to the PnP problem,” in Proceedings of the IEEE 11th International Conference on Computer Vision, 2007 (IEEE, 2007), pp. 1–8.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Pose priors for simultaneously solving alignment and correspondence,” in Proceedings of the 10th European Conference on Computer Vision: Part II (Springer-Verlag, 2008), pp. 405–418.

Navab, N.

N. Navab and O. Faugeras, “Monocular pose determination from lines: critical sets and maximum number of solutions,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (IEEE, 1993), pp. 254–260.

Negahdaripour, S.

B. K. P. Horn, H. Hilden, and S. Negahdaripour, “Closed-form solution of absolute orientation using orthonormal matrices,” J. Opt. Soc. Am. 5, 1127–1135 (1988).
[CrossRef]

Phong, T. Q.

T. Q. Phong, R. Horaud, A. Yassine, and P. D. Tao, “Object pose from 2D to 3D point and line correspondences,” Int. J. Comput. Vis. 15, 225–243 (1995).
[CrossRef]

Quan, L.

L. Quan and Z. Lan, “Linear N-point camera pose determination,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 774–780 (1999).
[CrossRef]

Richetin, M.

M. Dhome, M. Richetin, J. thierry Lapreste, and G. Rives, “Determination of the attitude of 3-D objects from a single perspective view,” IEEE Trans. Pattern Anal. Mach. Intell. 11, 1265–1278 (1989).
[CrossRef]

Riseman, E. M.

J. R. Beveridge and E. M. Riseman, “Optimal geometric model matching under full 3d perspective,” Comput. Vis. Image Understand. 61, 351–364 (1995).
[CrossRef]

Rives, G.

M. Dhome, M. Richetin, J. thierry Lapreste, and G. Rives, “Determination of the attitude of 3-D objects from a single perspective view,” IEEE Trans. Pattern Anal. Mach. Intell. 11, 1265–1278 (1989).
[CrossRef]

Samet, H.

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “SoftPOSIT: Simultaneous pose and correspondence determination,” Int. J. Comput. Vis. 59, 259–284 (2004).
[CrossRef]

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “Simultaneous pose and correspondence determination using line features,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003 (IEEE, 2003), pp. II-424–II-431.

Stricker, D.

H. Wuest, F. Vial, and D. Stricker, “Adaptive line tracking with multiple hypotheses for augmented reality,” in Proceedings of the 4th IEEE/ACM International Symposium on Mixed and Augmented Reality (IEEE, 2005), pp. 62–69.

Tao, P. D.

T. Q. Phong, R. Horaud, A. Yassine, and P. D. Tao, “Object pose from 2D to 3D point and line correspondences,” Int. J. Comput. Vis. 15, 225–243 (1995).
[CrossRef]

Tepichin, E.

thierry Lapreste, J.

M. Dhome, M. Richetin, J. thierry Lapreste, and G. Rives, “Determination of the attitude of 3-D objects from a single perspective view,” IEEE Trans. Pattern Anal. Mach. Intell. 11, 1265–1278 (1989).
[CrossRef]

Umeyama, S.

S. Umeyama, “Least-squares estimation of transformatio parameters between two point patterns,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 376–380 (1991).
[CrossRef]

Vaidya, V. G.

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

Vial, F.

H. Wuest, F. Vial, and D. Stricker, “Adaptive line tracking with multiple hypotheses for augmented reality,” in Proceedings of the 4th IEEE/ACM International Symposium on Mixed and Augmented Reality (IEEE, 2005), pp. 62–69.

Wolfson, H. J.

Y. Lamdan and H. J. Wolfson, “Geometric hashing: a general and efficient model-based recognition scheme,” in Proceedings of the Second International Conference on Computer Vision (IEEE, 1988), pp. 238–249.

Wuest, H.

H. Wuest, F. Vial, and D. Stricker, “Adaptive line tracking with multiple hypotheses for augmented reality,” in Proceedings of the 4th IEEE/ACM International Symposium on Mixed and Augmented Reality (IEEE, 2005), pp. 62–69.

Wunsch, P.

P. Wunsch and G. Hirzinger, “Registration of cad models to images by iterative inverse perspective matching,” in Proceedings of the 13th International Conference on Pattern Recognition (IEEE, 1996), pp. 78–83.

Yassine, A.

T. Q. Phong, R. Horaud, A. Yassine, and P. D. Tao, “Object pose from 2D to 3D point and line correspondences,” Int. J. Comput. Vis. 15, 225–243 (1995).
[CrossRef]

Zhuang, X.

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

Appl. Opt.

Commun. ACM

M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM 24, 381–395 (1981).
[CrossRef]

Comput. Vis. Graph. Image Process.

Y. Liu, T. S. Huang, and O. D. Faugeras, “A linear algorithm for motion estimation using straight line correspondences,” Comput. Vis. Graph. Image Process. 44, 35–57 (1988).
[CrossRef]

R. Kumar and A. R. Hanson, “Robust methods for estimating pose and a sensitivity analysis,” Comput. Vis. Graph. Image Process. 60, 313–342 (1994).
[CrossRef]

Comput. Vis. Image Understand.

J. R. Beveridge and E. M. Riseman, “Optimal geometric model matching under full 3d perspective,” Comput. Vis. Image Understand. 61, 351–364 (1995).
[CrossRef]

F. Jurie, “Solution of the simultaneous pose and correspondence problem using gaussian error model,” Comput. Vis. Image Understand. 73, 357–373 (1999).
[CrossRef]

S. Christy and R. Horaud, “Iterative pose computation from line correspondences,” Comput. Vis. Image Understand. 73, 137–144 (1999).
[CrossRef]

IEEE Trans. Pattern Anal. Mach. Intell.

C.-P. Lu, G. D. Hager, and E. Mjolsness, “Fast and globally convergent pose estimation from video images,” IEEE Trans. Pattern Anal. Mach. Intell. 22, 610–622 (2000).
[CrossRef]

A. Ansar and K. Daniilidis, “Linear pose estimation from points or lines,” IEEE Trans. Pattern Anal. Mach. Intell. 25, 578–589 (2003).
[CrossRef]

M. Dhome, M. Richetin, J. thierry Lapreste, and G. Rives, “Determination of the attitude of 3-D objects from a single perspective view,” IEEE Trans. Pattern Anal. Mach. Intell. 11, 1265–1278 (1989).
[CrossRef]

J. S. Beis and D. G. Lowe, “Indexing without invariants in 3D object recognition,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 1000–1015 (1999).
[CrossRef]

S. Umeyama, “Least-squares estimation of transformatio parameters between two point patterns,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 376–380 (1991).
[CrossRef]

Y. Liu, T. S. Huang, and L. D. Faugeras, “Determination of camera location from 2-D to 3-D line and point correspondences,” IEEE Trans. Pattern Anal. Mach. Intell. 12, 28–37 (1990).
[CrossRef]

L. Quan and Z. Lan, “Linear N-point camera pose determination,” IEEE Trans. Pattern Anal. Mach. Intell. 21, 774–780 (1999).
[CrossRef]

H. H. Chen, “Pose determination from line-to-plane correspondences: existence condition and closed-form solutions,” IEEE Trans. Pattern Anal. Mach. Intell. 13, 530–541 (1991).
[CrossRef]

IEEE Trans. Syst. Man Cybern.

R. M. Haralick, H. Joo, C.-N. Lee, X. Zhuang, V. G. Vaidya, and M. B. Kim, “Pose estimation from corresponding point data,” IEEE Trans. Syst. Man Cybern. 19, 1426–1446 (1989).
[CrossRef]

Image Vis. Comput.

C.-N. Lee and R. M. Haralick, “Statistical estimation for exterior orientation from line-to-line correspondences,” Image Vis. Comput. 14, 379–388 (1996).
[CrossRef]

Int. J. Comput. Vis.

T. Q. Phong, R. Horaud, A. Yassine, and P. D. Tao, “Object pose from 2D to 3D point and line correspondences,” Int. J. Comput. Vis. 15, 225–243 (1995).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, S. Christy, “Object pose: the link between weak perspective, paraperspective, and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

W. E. L. Grimson, “Object recognition by computer: the role of geometric constraints,” Int. J. Comput. Vis. 31, 350–504 (1990).

D. DeMenthon and L. Davis, “Model-based object pose in 25 lines of code,” Int. J. Comput. Vis. 15, 123–141 (1995).
[CrossRef]

R. Horaud, F. Dornaika, B. Lamiroy, and S. Christy, “Object pose: the link between weak perspective, paraperspective and full perspective,” Int. J. Comput. Vis. 22, 173–189 (1997).
[CrossRef]

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “SoftPOSIT: Simultaneous pose and correspondence determination,” Int. J. Comput. Vis. 59, 259–284 (2004).
[CrossRef]

J. Opt. Soc. Am.

B. K. P. Horn, H. Hilden, and S. Negahdaripour, “Closed-form solution of absolute orientation using orthonormal matrices,” J. Opt. Soc. Am. 5, 1127–1135 (1988).
[CrossRef]

Other

H. Wuest, F. Vial, and D. Stricker, “Adaptive line tracking with multiple hypotheses for augmented reality,” in Proceedings of the 4th IEEE/ACM International Symposium on Mixed and Augmented Reality (IEEE, 2005), pp. 62–69.

N. Navab and O. Faugeras, “Monocular pose determination from lines: critical sets and maximum number of solutions,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (IEEE, 1993), pp. 254–260.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Accurate non-iterative O(n) solution to the PnP problem,” in Proceedings of the IEEE 11th International Conference on Computer Vision, 2007 (IEEE, 2007), pp. 1–8.

Y. Lamdan and H. J. Wolfson, “Geometric hashing: a general and efficient model-based recognition scheme,” in Proceedings of the Second International Conference on Computer Vision (IEEE, 1988), pp. 238–249.

P. Wunsch and G. Hirzinger, “Registration of cad models to images by iterative inverse perspective matching,” in Proceedings of the 13th International Conference on Pattern Recognition (IEEE, 1996), pp. 78–83.

P. David, D. Dementhon, R. Duraiswami, and H. Samet, “Simultaneous pose and correspondence determination using line features,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003 (IEEE, 2003), pp. II-424–II-431.

F. Moreno-Noguer, V. Lepetit, and P. Fua, “Pose priors for simultaneously solving alignment and correspondence,” in Proceedings of the 10th European Conference on Computer Vision: Part II (Springer-Verlag, 2008), pp. 405–418.

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (13)

Fig. 1.
Fig. 1.

Geometry of the camera model.

Fig. 2.
Fig. 2.

Object-space error of line correspondence.

Fig. 3.
Fig. 3.

(a) Example of the simulation of the 3D object lines and (b) the detected infinite 2D image lines. The 3D lines are generated in 3D space by connecting two random endpoints that are not too close. Their projections are drawn in either red or pinkish red. The red ones are those that have true detected 2D lines, while the pinkish red ones are those have no true detections that may caused by occlusions. The detected 2D image lines are either in blue or green. Blue denotes nonclutter, while green means clutter detections. Gaussian noise is added to perturb the locations of the endpoints of each 2D line. Here the noise deviation σ=2 pixels.

Fig. 4.
Fig. 4.

Example of using our simultaneous pose and correspondence method on a synthetic cluttered image. In all simulated images (image size is 640×480), blue lines are the 2D detected lines, among which there are some clutter lines; red lines are the projections of all the 3D lines at the current estimated pose. (a) True pose, (b)–(f) successive pose update steps. The correct pose is found by five update steps. The experimental parameters: N=20, po=0.2, pc=0.4, and σ=2.

Fig. 5.
Fig. 5.

Convergence percentage for different combinations of settings of occlusion po and clutter pc. The number of object lines varies from 20 to 60.

Fig. 6.
Fig. 6.

Computational time as a function of the number of 3D object lines. The occlusion percentage po=0.4, and the clutter percentage pc=0.6.

Fig. 7.
Fig. 7.

Relative rotation and translation error as a function of image noise when the number of lines is fixed to be 8.

Fig. 8.
Fig. 8.

Relative rotation and translation error as a function of the number of object lines when the standard deviation of image noise is fixed to be 3 pixels.

Fig. 9.
Fig. 9.

(a) Percentage of convergence when initial poses are generated from a multinormal distribution, (b) the number of iterations, and (c) the running time as a function of the number of lines.

Fig. 10.
Fig. 10.

Relative rotation and translation error as a function of the percentage of outliers. For all trials, the number of object lines is 25 and the standard deviation of image noise is 3.

Fig. 11.
Fig. 11.

(a) 3D object model; (b) 2D lines are extracted by Hough transform; (c) a number of clutter lines or repeat detections are removed to limit the number of valid lines to between 50 and 80 (green lines are the final detections). Note that although the lines are detected as segments, they are used as infinity lines in all algorithms.

Fig. 12.
Fig. 12.

(a) Example of the mean pose of one prior Gaussian component, (b)–(d) example results for different occlusion cases.

Fig. 13.
Fig. 13.

Pose estimation for 3D object tracking on a sequence recorded with a moving camera.

Equations (34)

Equations on this page are rendered with MathJax. Learn more.

niT=miTA,
niTRdi=0,
niT(RPi+t)=0.
li=fl(ui1,ui2)=(v2v1u1u2,v1u2v2u1u1u2)T,
Σi=J(ui1,ui2)Σi*(J(ui1,ui2))T,
Σij=J(Pij)Σgp(J(Pij))T,
(lilj)T(Σi)1(lilj)M2
f(lj,pg)=[mjARdimjA(Rpi+t)].
pg*=pgQf(lj,pg),
Σg*p=(IQM)Σgp,
Q=ΣgpMT(U+MΣgpMT)1,M=fpg,U=fljU*(flj)T,
E(p)=i=1Keij+(MK)T,
eij=w1njTRdi2+w2njT(RPi+t)2.
k=argjmineij.
E(R,t)=i=1N(niTRdi)2+i=1N(niT(RPi+t))2.
Rdi=KiRdi.
RPi+t=Ki(RPi+t).
eid=(IKi)Rdi,
eiP=(IKi)(RPi+t).
E1(R)=i=1Neid2=i=1N(IKi)Rdi2,
E2(R,t)=i=1NeiP2=i=1N(IKi)(RPi+t)2.
KiKix,xR3,
KiT=Ki,
Ki2=KiKiT=Ki.
f(R)=ARB2,
R=USVT.
S={Iifdet(ABT)0,diag(1,1,,1,1)else.
S={Iifdet(U)det(V)=1,diag(1,1,,1,1)ifdet(U)det(V)=1.
A=(K1Rd1,K2Rd2,,KNRdN),B=(d1,d2,,dN).
E1(R)=ARB2.
t=t(R)=(i=1N(IKi))1i=1N(KiI)RPi.
C=[a1a2aNb1b2bNc1c2cN]T.
E(R,t)=iN(IVi)(RPi+t)2,
Rk+1=argminRi=1NRPi+tViqik2,

Metrics