대한전자공학회논문지SP (Journal of the Institute of Electronics Engineers of Korea SP)

대한전자공학회 (The Institute of Electronics and Information Engineers)
권호 표지

겹쳐진 3차원 물체의 2차원 영상에서 가리는 물체의 구분기법

Separation of the Occluding Object from the Stack of 3D Objects Using a 2D Image

  • 송필재 (동서울대학 컴퓨터시스템과) ;
  • 홍민철 (숭실대학교 정보통신전자공학) ;
  • 한헌수 (숭실대학교 정보통신전자공학부)
  • 발행 : 2004.03.01
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

겹쳐진 물체의 분리에 대한 기존의 연구들은 주로 모델기반 정합방법을 사용하고 있어서 적용가능물체가 2차원물체에 한정되었고, 모델물체의 수가 증가하면 정합과정에 소요되는 시간이 지수적으로 증가하는 문제를 갖는다. 이러한 문제를 해결하기 위해 본 논문에서는 곡면을 포함하는 3차원물체를 대상으로 물체에 대한 정보가 없는 상황에서도 가리는 물체를 찾아내는 알고리즘을 제시한다. 이 방법을 이용함으로써 겹쳐진 물체의 인식이 독립된 하나의 물체를 인식하는 문제로 단순화될 수 있다. 제안하는 알고리즘은 물체를 면의 결합으로 해석하고 면들은 경계선을 속성으로 표현된다. 3차원 물체의 겹침은 2차원 영상에서 면의 겹침으로 보여 지고 면의 겹침은 경계선의 교차로 나타나는 특성을 이용하여 겹침의 형태를 경계선의 형태로 일반화하여 분류하는 기법을 사용하였다. 면사이의 겹침 관계를 면특징 관계도를 이용하여 표현하기 위해 관계계수를 정의하였고 관계계수의 값은 겹침의 형태를 보여주도록 개념화하였다. 제안된 알고리즘의 성능은 산업현장에서 사용되는 표준 부품을 임의로 겹치게하여 영상을 획득한 후 가리는 물체를 구분하는 실험을 통해 입증하였다.

Conventional algorithms of separating overlapped objects are mostly based on template matching methods and thus their application domain is restricted to 2D objects and the processing time increases when the number of templates (object models) does. To solve these problems, this paper proposes a new approach of separating the occluding object from the stack of 3D objects using the relationship between surfaces without any information on the objects. The proposed algorithm considers an object as a combination of surfaces which are consisted with a set of boundary edges. Overlap of 3D objects appears as overlap of surfaces and thus as crossings of edges in 2D image. Based on this observation, the types of edge crossings are classified from which the types of overlap of 3D objects can be identified. The relationships between surfaces are represented by an attributed graph where the types of overlaps are represented by relation values. Using the relation values, the surfaces pertained to the same object are discerned and the overlapping object on the top of the stack can be separated. The performance of the proposed algorithm has been proved by the experiments using the overlapped images of 3D objects selected among the standard industrial parts.

키워드

참고문헌

  1. Kawaguchi T., Nagao M., 'Recognition of occluded objects by a genetic algorithm', Proceedings of 14th International Conference on Pattern Recognition, Vol. 1, 16-20 Aug. 1998, pp. 233-237 https://doi.org/10.1109/ICPR.1998.711124
  2. Dockstader S.L., Tekalp A.M., 'Tracking multiple objects in the presence of articulated and occluded motion', Proceedings of 2000 Workshop on Human Motion, 7-8 Dec. 2000, pp. 88-95 https://doi.org/10.1109/HUMO.2000.897376
  3. Kimachi M., Wu Y., Ogata S., 'A vehicle recognition method robust against vehicles' overlapping based on stereo vision', Proceedings of 1999 IEEE/IEEJ/JSAI International Conference on Intelligent Transportation Systems, 5-8 Oct. 1999, pp. 865-869 https://doi.org/10.1109/ITSC.1999.821176
  4. Mayez Al-Mouhamed, 'ARobust Gross-to-Fine Pattern Recognition System', Industrial Electronics, IEEE Transactions on , Volume: 48 Issue: 6, Dec. 2001, pp. 1226-1237 https://doi.org/10.1109/41.969403
  5. M.S. Abou-El-Ela, H. El-Amroussy, 'A Machine Vision System For The Recognition And Positioning Of Two-Dimensional Partially Occluded Objects', IEEE, 1996 https://doi.org/10.1109/MELCON.1996.551397
  6. Yuichi Nakamura, Makoto Nagao, 'Recognition of Overlapping 2-D Objects by Local Feature Construction Method', IEEE, 1988 https://doi.org/10.1109/ICPR.1988.28436
  7. R. A. Brooks,'Symbolic reasoning among 3D models and 2D images,' Artificial Intelligence, vol. 17, 1981, pp. 285-348 https://doi.org/10.1016/0004-3702(81)90028-X
  8. Wen-Jing Li, Tong Lee, 'Object recognition by sub-scene graph matching', IEEE International Conference on Robotics and Automation, Vol. 2, 2000, pp. 1459-1464 https://doi.org/10.1109/ROBOT.2000.844803
  9. R.Cucchiara, E.Lamma, P.Mello, M.Milano, M.Piccardi, '3D Object Recognition by VC-graphs and Interactive Constraint Satisfaction', Proceedings of International Conference on Image Analysis and Processing, pp. 508-513, 1999.1999 https://doi.org/10.1109/ICIAP.1999.797646
  10. Umberto Castellani, Salvatore Livatino, Robert B. Fisherge,'Improving Environment Modelling by Edge Occlusion Surface Completion', 3D Data Processing Visualization and Transmission, 2002. Proceedings. First International Symposium on, 19-21 June 2002, pp. 672-675 https://doi.org/10.1109/TDPVT.2002.1024138
  11. Rachid Benlamri, 'Curved Shapes Construction for ZObject Recognition', Geometric Modeling and Processing, 2002. Proceedings, 10-12 July 2002, pp. 197-204 https://doi.org/10.1109/GMAP.2002.1027511
  12. Stan Z. Li, 'Recognizing Multiple Overlapping Objects in Image: An Optimal Formulation', IEEE Transactions on Image Processing, Vol. 9, No 2, February 2000, pp. 273-277 https://doi.org/10.1109/83.821741
  13. 송필재, 최홍주, 차형태, 한헌수, 'Face Relation Features for Separating Overlapped Objects in a 2D Image', 대한전자공학회 논문지, 제38권 SP편 제1호, pp. 54-68