한국산학기술학회논문지 (Journal of the Korea Academia-Industrial cooperation Society)

  • 제20권5호
  • /
  • Pages.530-537
  • /
  • 2019
  • /
  • 1975-4701(pISSN)
  • /
  • 2288-4688(eISSN)
한국산학기술학회 (The Korea Academia-Industrial cooperation Society)
권호 표지
DOI QR코드

DOI QR Code

시스템 함수 및 변복조 개념 적용 능동 방식 3차원 물체 좌표 복원

Concepts of System Function and Modulation-Demodulation based Reconstruction of a 3D Object Coordinates using Active Method

  • 이덕우 (계명대학교 공과대학 컴퓨터공학전공) ;
  • 김지수 (계명대학교 공과대학 컴퓨터공학전공) ;
  • 박철형 (계명대학교 공과대학 컴퓨터공학전공)
  • Lee, Deokwoo (Department of Computer Engineering, Keimyung University) ;
  • Kim, Jisu (Department of Computer Engineering, Keimyung University) ;
  • Park, Cheolhyeong (Department of Computer Engineering, Keimyung University)
  • 투고 : 2019.02.08
  • 심사 : 2019.05.03
  • 발행 : 2019.05.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

본 논문에서는 시스템함수 및 변복조의 개념을 3차원 복원 문제에 적용하는 알고리즘을 제안한다. 시스템의 유일한 특성을 정의하는 시스템 함수 (또는 시스템 응답)를를 일반적인 신호처리 또는 제어시스템에서 결정하듯이, 본 논문에서는 적절한 입력과 출력신호를 선택한 다음 3차원 물체의 특성을 결정짓는 시스템 함수를 결정한다. 본 논문에서는 3차원 복원 문제를 두 가지 방법의 시스템 함수 문제로 풀어 나간다. 첫 번째 방법은 입력과 출력 신호를 각각 3차원 물체의 면에 투영된 원형 빛 패턴과 카메라(2차원 이미지 면)가 획득한 패턴이 투영된 3차원 물체의 이미지로 정의하여 3차원 물체의 특성을 나타내는 시스템 함수를 정의 하는 것이다. 두 번째 방법은 입력과 출력 신호를 각각 복원되어야 할 3차원 물체의 좌표와 카메라가 획득한 빛 패턴이 투영된 3차원 물체의 이미지로 정의하여 입력 신호를 추정하는 문제로 해석하는 것이다. 첫 번째 방법은 일반적인 입출력 함수의 비(ratio)로부터 시스템 함수를 구하는 것이고 두 번째 방법은 신호의 변조와 복조 과정으로부터 원래의 전송된 신호 (입력) 를 추정하는 것처럼 입력 신호인 3차원 물체의 좌표를 추정하는 것이다.

In this paper we propose a novel approach to representation of the 3D reconstruction problem by employing a concept of system function that is defined as the ratio of the output to the input signal. Akin to determination of system function (or system response), this paper determines system function by choosing (or defining) appropriate input and output signals. In other words, the 3D reconstruction using structured circular light patterns is reformulated as determination of system function from input and output signals. This paper introduces two algorithms for the reconstruction. The one defines the input and output signals as projected circular light patterns and the images overlaid with the patterns and captured by camera, respectively. The other one defines input and output signals as 3D coordinates of the object surface and the image captured by camera. The first one leads to the problem as identifying the system function and the second one leads to the problem as estimation of an input signal employing concept of modulation-demodulation theory. This paper substantiate the proposed approach by providing experimental results.

키워드

SHGSCZ_2019_v20n5_530_f0001.png 이미지

Fig. 1. System function establishes a relationship between input and output signal.

SHGSCZ_2019_v20n5_530_f0002.png 이미지

Fig. 2. Transmitted signal is estimated by demodulation of the received signal that is modulation of the transmitted one.

SHGSCZ_2019_v20n5_530_f0003.png 이미지

Fig. 3. Flowchart of a 3D reconstruction using structured circular light patterns

SHGSCZ_2019_v20n5_530_f0004.png 이미지

Fig. 5. Based on the Thales’ Theorem, and on the properties of a circular pattern, relationship between M and m is established.

SHGSCZ_2019_v20n5_530_f0005.png 이미지

Fig. 6. A variable t represents the positions of circular patterns projected onto the 3D object.

SHGSCZ_2019_v20n5_530_f0006.png 이미지

Fig. 7. Modulation and demodulation in communication systems

SHGSCZ_2019_v20n5_530_f0007.png 이미지

Fig. 8. Estimation of 3D coordinates of an object employing the concept of modulation and demodulation

SHGSCZ_2019_v20n5_530_f0008.png 이미지

Fig. 9. Input A : Original circular pattern, Ouput B : Deformed pattern by the object surface, System H : Reconstructed 3D coordinates (depth) of the object.

SHGSCZ_2019_v20n5_530_f0009.png 이미지

Fig. 10. 3D reconstruction of the object overlaid with the light patterns based on MODEM reconstruction.

SHGSCZ_2019_v20n5_530_f0010.png 이미지

Fig. 11. Quantifying reconstruction result with the number of circular light patterns

Fig. 4. Choosing input, output and system function in 3D measurement setup

SHGSCZ_2019_v20n5_530_t0001.png 이미지

참고문헌

  1. D. Scharstein, R. Szeliski, "High-accuracy stereo depth maps using structured light", Proceedings of IEEE Computer Society on Computer Vision and Pattern Recognition, IEEE, Madison, WI, USA, June 2003. DOI: https://doi.org/10.1109/CVPR.2003.1211354
  2. J. Salvi, J. Pages, J. Batlle, "Pattern codification strategies in structured light systems", Pattern Recognition, Vol. 37, No. 4, pp.827-849, April 2004. DOI: https://doi.org/doi.org/10.1016/j.patcog.2003.10.002
  3. J. Geng, "Structured-light 3D surface imaging: a tutorial", Advances in Optics and Photonics, Vol. 3, No. 2, pp.128-160, March 2011. DOI: https://doi.org/doi.org/10.1364/AOP.3.000128
  4. D. Lee, H. Krim, "A Sampling Theorem for a 2D Surface", Proceedings of International Conference on Scale Space and Variational Methods in Computer Vision , pp.556-567, Springer, Berlin, Heidelberg, Ein-Gedi, Israel, May, 2011. DOI: https://doi.org/doi.org/10.1007/978-3-642-24785-9_47
  5. A. Papoulis, Signal Analysis, 1st Ed., Mcgraw-Hill College, 1977. ISBN : 978-0070484603
  6. D.L. Donoho, "Compressed Sensing", IEEE Transactions on Information Theory, Vol. 52, No. 4, pp.1289-1306, April 2006. DOI: https://doi.org/10.1109/TIT.2006.871582
  7. A.V. Oppenheim, A.S. Willsky, S. Hamid, Signals ans Systems, 2nd Ed., Peasrson, 1996.
  8. S. Haykin, M. Moher, Communication Systems, 5th Ed., Wiley, 2009.
  9. D. Lee, H. Krim, "3D Surface Reconstruction Using Structured Circular Light Patterns", Proceedings of International Conference on Advanced Concepts for Intelligent Vision Systems , pp. 279-289, Springer, Berlin, Heidelberg, Sydney, Australia, December 2010. DOI: https://doi.org/doi.org/10.1007/978-3-642-17688-3_27
  10. A. Dipanda, S. Woo, "Towards a real-time 3D shape reconstruction using a structured light system", Pattern Recognition , Vol. 38, No. 10, pp.1632-1650, October 2005. DOI: https://doi.org/10.1016/j.patcog.2005.01.006
  11. H. Kawasaki, R. Furukawa, R. Sagawa, Y. Yagi, "Dynamic scene shape reconstruction using a single structured light pattern", Proceedings of IEEE Computer Society on Computer Vision and Pattern Recognition , IEEE, Anchorage, AK, USA, pp. 1-8, June. 2008. DOI : 10.1109/CVPR.2008.4587702
  12. J. Pages, J. Salvi, R. Garcia, C. Matabosch, "Overview of coded light projection techniques for automatic 3D profiling", Proceedings of IEEE International Conference on Robotics and Automation, IEEE, Taipei, Taiwan, pp. 133-138, September 2003. DOI: https://doi.org/10.1109/ROBOT.2003.1241585
  13. J. Salvi, J. Batlle, E. Mouaddib, "A robust-coded pattern projection for dynamic 3D scene measurement", Pattern Recognition Letters , Vol. 19, No. 11, pp.1055-1065, September 1998. DOI: https://doi.org/10.1016/S0167-8655(98)00085-3
  14. O. Faugeras and Q. T. Luong, The Geometry of Multiple Images, The MIT Press 2001. ISBN : 0262062208
  15. T.L.N. Nguyen, H. Jung, Y. Shin, "A Signal Detection and Estimation Method Based on Compressive Sensing", The Journal of Korean Institute of Communication and Information Science , Vol. 40, No. 6, pp.1024-1031, June 2015. DOI: https://doi.org/10.7840/kics.2015.40.6.1024
  16. S. Jeong, D. Lim, "A Study on the Reconstruction of a Frame Based Speech Signal through Dictionary Learning and Adaptive Compressed Sensing", The Journal of Korean Institute of Communication and Information Science , Vol. 37, No. 12, pp.1122-1132, December 2012. DOI: https://doi.org/10.7840/kics.2012.37A.12.1122
  17. D. Kang, H. Kim, K. Park, W. Oh, "Parameter Derivation for Reducing ISI in 2-Dimensional Faster-than-Nyquist Transmission", The Journal of Korean Institute of Communication and Information Science, Vol. 41, No. 10, pp.1147-1154, October 2016. DOI: https://doi.org/10.7840/kics.2016.41.10.1147
  18. B. Sklar, Digital Communications: Fundamentals and Applications, 2nd Ed. Prentice Hall 2001.