Journal of the Korea Institute of Information and Communication Engineering (한국정보통신학회논문지)

The Korea Institute of Information and Commucation Engineering (한국정보통신학회)
권호 표지
DOI QR코드

DOI QR Code

Parameter analysis in Fast Global Registration to improve accuracy and speed

고속 전역 정합법에서 정밀도 및 속도 향상을 위한 매개변수 분석

  • Received : 2021.04.07
  • Accepted : 2021.04.15
  • Published : 2021.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

The transforming process of point clouds with its local coordinates into a global coordinate is called registration. In contrast to the local registration which takes a long time to calculate and performs precision registration after initial rough positioning, the global registration calculates the corresponding points for registration and performs at once, so it is generally faster than the local registration, and can perform it regardless of the initial position. Among the global methods, the Fast Global Registration is one of the widely used methods due to its fast performance. However, lots of parameters should be set to increase the registration accuracy and speed. In this paper, after analyzing and experimenting the parameters and propose parameters that work effectively in actual registration. The proposed result will be helpful in setting the direction when it is necessary to use the Fast Global Registration method.

정합은 고유 좌표를 가지고 있는 점군을 전역 좌표로 변환하는 과정이다. 지역 정합은 계산 시간이 오래 걸리고 대략적인 위치를 맞춘 후 정밀 정합을 수행하고, 전역 정합은 정합에 이용할 대응점을 계산하고 한 번에 정합하기 때문에 일반적으로 지역 정합법에 비해 속도가 빠르고, 초기 위치에도 상관이 없다. 전역 정합 방법 중 고속 전역 정합법은 성능이 우수하여 많이 사용하는 방법 중 하나이다. 하지만 정합 정밀도와 속도를 높이기 위해서는 많은 매개변수가 필요하다. 본 논문에서는 이와 같은 매개변수들을 분석하고 실험하여 실제 정합 시 유효하게 작용하는 매개변수를 제안한다. 제안한 결과는 고속 전역 정합법을 활용해야 하는 경우 방향 설정에 도움이 될 것이다.

Keywords

References

  1. S. Lim, "Effective criterion for evaluating registration accuracy," Journal of the Korea Institute of Information and Communication Engineering, vol. 25, no. 5, pp. 652-658, 2021. https://doi.org/10.6109/JKIICE.2021.25.5.652
  2. K. Kwon, "A weighted points registration method to analyze dimensional errors occurring during shipbuilding process," Transactions of the Society of CAD/CAM Engineers, vol. 21, no. 2, pp. 151-158, 2016. https://doi.org/10.7315/CADCAM.2016.151
  3. Z. H. Nejad and M. Nasri, "An adaptive image registration method based on SIFT features and RANSAC transform," Computers & Electrical Engineering, vol. 62, pp. 524-537, 2017. https://doi.org/10.1016/j.compeleceng.2016.11.034
  4. M. He, L. Huang, B. Zhao, B. Chen, and B. Hu, "Advanced functional materials in solid phase extraction for ICP-MS determination of trace elements and their species - A review," Analytica Chimica Acta, vol. 973, no. 22 pp. 1-24, 2017. https://doi.org/10.1016/j.aca.2017.03.047
  5. Z. Wu, H. Chen, S. Du, M. Fu, N. Zhou, and N. Zheng, "Correntropy based scale ICP algorithm for robust point set registration," Pattern Recognition, vol. 93, pp. 14-24, 2019. https://doi.org/10.1016/j.patcog.2019.03.013
  6. J. Yang, H. Li, D. Campbell, and Y. Jia, "Go-ICP: A globally optimal solution to 3D ICP pointset registration," in Proceedings IEEE/CVF International Conference on Computer Vision, 2016.
  7. Q. Y. Zhou, P. Jaesik, and K. Vladlen, "Fast global registration," in Proceedings European Conference on Computer Vision, Netherlands, 2016.
  8. Open3D project [Internet]. Available: http://www.open3d.org/.
  9. R. B. Rusu, N. Blodow, and M. Beetz, "Fast point feature histograms (FPFH) for 3D registration," in Proceedings IEEE International Conference on Robotics and Automation, 2009.
  10. J. T. Barron, "A general and adaptive robust loss function," in Proceeding of Computer Vision and Pattern Recognition, USA, pp. 4331-4339, 2019.
  11. Stanford University 3D Scan Repository [Internet]. Available: http://graphics.stanford.edu/data/.