A Study on Disparity Correction of Occlusion using Occluding Patterns

가려짐 패턴을 이용한 가려짐 영역의 시차 교정에 관한 연구

  • Kim Dae-Hyun (Digital Media R'D Center, Samsung Electronic Co., LTD) ;
  • Choi Jong-Soo (Dept. Image Eng., Graduate School of AIM, Chung-Ang Univ)
  • 김대현 (삼성전자 DM 연구소) ;
  • 최종수 (중앙대학교 참단영상대학원 영상공학과)
  • Published : 2005.07.01

Abstract

In this paper, we propose new smoothing filters, i.e., occluding patterns that can accurately correct disparities of occluded areas in the estimated disparity map. An image is composed of several layers and each layer presents similar disparity. Furthermore, the distribution of the estimated disparities has a specific direction around the boundary of the occlusion, and this distribution presents the different direction with respect to the left- and the right-based disparity map. However, typical smoothing filters, such as mean filter and median filter, did not take into account those characteristic. So, they can decrease some error, but they cannot guarantee the accuracy of the corrected disparity. On the contrary, occluding patterns can accurately correct disparities of occluded areas because they consider both the characteristic that occlusion occurs and the characteristic that disparities of the occlusion are ranged, from estimated disparity maps with respect to the left and the right images. We made experiments on occluding patterns with some real stereo image set, and as a result, we can correct disparities of occluded areas more accurately than typical smoothing filters did.

본 논문에서 우리는 추정된 시차지도에서 가려짐 영역의 시차를 교정하는 새로운 스무딩 필터인 가려짐 패턴 (occluding patterns)을 제안한다. 영상은 몇 개의 계층으로 구성되어 있고, 각각의 계층은 유사한 시차를 나타낸다. 그리고 추정된 시차들은 가려짐 영역의 경계 주변에서 특정한 방향성을 갖고 분포하며, 이러한 시차 분포의 방향성은 좌우 시차지도에 대해서 서로반대이다. 그러나 평균값 필터 또는 중간값 필터와 같은 기존의 스무딩 필터는 이러한 시차의 분포 특성을 고려하지 않고 스무딩을 수행하기 때문에 오차는 줄일 수 있으나, 교정된 시차의 정확성은 보장되지 않았다. 이와 반대로, 본 논문에서 제안하는 가려짐 패턴은 좌우 영상에서 각각 추정된 시차지도에 대해서 가려짐 영역이 발생하는 특성과 가려짐 영역에서의 시차 분포 특성을 함께 고려하여 정확하게 가려짐 영역의 시차를 교정한다. 본 논문에서 제안한 가려짐 패턴은 다양한 실험 영상에 적용하여 실험하였고, 그 결과 기존의 스무딩 방법에 비해서 정확하게 시차를 교정하는 것을 확인하였다.

Keywords

References

  1. H.H. Baker and T.O. Binford, 'Depth from edge and intensity based stereo,' Proc. 7th Int'l Joint Conf. Artificial Intelligence, pp.631-636, 1981
  2. Y.Ohta and T.Kanade, 'Stereo by intra- and inter-scanline search using dynamic programming,' IEEE Trans. Pattern Analysis and Machine Intelligence, Vol.7, pp.139-154, 1985 https://doi.org/10.1109/TPAMI.1985.4767639
  3. S.A. Lloyd, E.R. Haddow, and J.F. Boyce, 'A parallel binocular stereo algorithm utilizing dynamic programming and relaxation labelling,' Computer Vision, Graphics, and Image Processing, Vol.39, pp.202-225, 1987 https://doi.org/10.1016/S0734-189X(87)80166-4
  4. Takeo Kanade, Masatoshi Okutomi, 'A Stereo Matching Algorithm with an Adaptive Window: Theory and Experiment,' IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 16, no. 9, Sept, 1994 https://doi.org/10.1109/34.310690
  5. S.T. Barnard and M.A. Fischler, 'Stereo vision,' encyclopedia of artificial intelligence, pp. 1083-1090, New York: John Wiley, 1987
  6. C. J. Tasi and A. K. Katsaggelos, ' Dense Disparity Estimation with a Divide-and-Conquer Disparity Space Image Technique,' IEEE, Trans. Multimedia, Vol.1, No.1, pp.18-29, March, 1999 https://doi.org/10.1109/6046.748168
  7. C. Lawrence Zitnick and T. Kanade, 'A Cooperative Algorithm for Stereo Matching and Occlusion Detection', IEEE Trans. Pattern Anal. Machine Intell., vol. 22, no. 7, pp. 1-10, 2000 https://doi.org/10.1109/34.865184
  8. S.Intille and A.Bobick, 'Incorporating intensity edges in the recovery of occlusion regions,' Proc. Int'l Conf. Pattern Recognition, Vol.1, pp.674-77, Oct. 1994 https://doi.org/10.1109/ICPR.1994.576397
  9. Y.Nakamura, TMatsuura, K.Satoh, and Y.Ohta, 'Occlusion detectable stereo-occlusion patterns in camera matrix,' Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp.371-78, 1996 https://doi.org/10.1109/CVPR.1996.517099
  10. P.Fua, 'A parallel stereo algorithm that produces dense depth maps and preserves image features,' Machine Vision and Applications, Vol.6, pp.35-49, 1993 https://doi.org/10.1007/BF01212430
  11. P.N. Belhumeur and D.A. Mumford, 'A bayesian treatment of the stereo correspondence problem using half-occluded regions,' Proc. IEEE Conf. Computer Vision and Pattern Recognition, 1992 https://doi.org/10.1109/CVPR.1992.223143
  12. D. Geiger, B. Ladendorf, and A. Yuille, 'Occlusions and Binocular Stereo,' Int'l J. Computer Vision, Vol.14, pp.211-226, 1995 https://doi.org/10.1007/BF01679683
  13. S. Intille and A. Bobick, 'Disparity-space images and large occlusion stereo,' Proc. European Conf. Computer Vision, Vol.801, pp.179-186, May 1994
  14. G. Egnal and R. P. Wildes, 'Detecting Binocular Half-Occlusion : Empirical Comparisons of Five Approaches', IEEE Trans. Pattern Anal. Machine Intell., vol. 24, no 8, pp. 1127-1133, August 2002 https://doi.org/10.1109/TPAMI.2002.1023808