SSOR을 이요한 강인한 F-행렬의 추정

The Robust Estimation of Fundamental Matrix Using the SSOR

  • 발행 : 2002.01.01

초록

시점을 달리하는 영상으로부터 3차원 장면 복구는 두 영상의 에피폴라 기하구조를 나타내는 F-행렬을 계산함으로서 가능하다. F-행렬을 계산하기 위해 입력으로 주어지는 두 영상의 일치점에는 잘못된 정합점과 같은 잡음을 포함하고 있기 때문에 정확한 F-행렬의 계산은 많은 오류를 가지게 된다. 따라서 본 논문에서는 에피폴라 기하구조에 영향을 미치는 잡음의 종류를 크게 outlier와 미세잡음으로 구분하였다. 상대적으로 에피폴라 기하구조에 영향을 크게 미치는 outlier를 단계적으로 제거시킴으로써 잡음 환경에서도 효과적으로 F-행렬을 계산할 수 있는 SSOR 알고리즘을 제안한다. 제안 알고리즘의 성능 평가를 위해 합성영상과 실 영상에서 실험하였으며 실험결과 제안 알고리즘이 기존의 알고리즘보다 성능이 우수함을 확인하였다.

Three-Dimensional scene reconstruction from images acquired with different viewpoints is possible as estimating Fundamental matrix(F-matrix) that indicates the epipolar geometry of two images. Correspondence points required to calculate F-matrix of two images include noise such as miss matches, so generally it is hard to calculate F-matrix accurately. In this paper, we classify noise into two types; outlier and minute noise. we propose SSOR algorithm that estimate F-matrix effectively. SSOR algorithm is rejecting outlier step by step in a noise environment. To evaluate the performance of proposed algorithm we simulated with synthetic images and real images. As a result of simulation we show that proposed algorithm is better than conventional algorithms.

키워드

참고문헌

  1. O. D. Faugeras, 'What can be seen in three dimensions with an uncalibrated stereo rig?,' Proccedings of the 2nd European Conference on Computer Vision, Santa Margherita, pp.563-578, 1992 https://doi.org/10.1007/3-540-55426-2_77
  2. P. Devemay and O. Faugeras, 'From projective to euclidean reconstruction,' Conference on Computer Vision and pattern Recogintion, IEEE Computer Soc. Press, pp. 264-269, 1996 https://doi.org/10.1109/CVPR.1996.517084
  3. R. Koch, M. Pollefeys and L. Van Gool, 'Automatic 3D Model Acquisition from Uncalibrated Image Sequences,' Proceedings Computer Graphics International, pp.567-604, Hannover, 1998 https://doi.org/10.1109/CGI.1998.694318
  4. H.C. Lonquet-Higgins, 'A Computer Algorithm for Reconstructing a Scene from Two Projections,' Nature 293, pp.133-135, Sep., 1981 https://doi.org/10.1038/293133a0
  5. R. I. Hartley, 'In Defence of the Eight-Point Algorithm,' Proceedings of IEEE International Conference on Computer Vision, pp.1064-1070, 1995 https://doi.org/10.1109/ICCV.1995.466816
  6. Boufama, B and Mohr, R, 'Epipole and fundamental matrix estimation using the virtual parallax peroperty,' In Proceedings ICCV95, pp.1030-1036, 1995 https://doi.org/10.1109/ICCV.1995.466821
  7. P. Torr and D. Murray, 'The Development and Comparison of Robust Methods for Estimating the Fundamental Matrix,' International Journal of Computer Vision, 3(24), pp.271-300, 1997 https://doi.org/10.1023/A:1007927408552
  8. J.Shi and C.Tomasi, 'Good features to track,' CVPR, pp.593-600, June, 1994 https://doi.org/10.1109/CVPR.1994.323794