УДК: 528.622.1:007.52

Robust point-cloud registration based on the maximum-likelihood method

Publication in Journal of Optical Technology

For Russian citation (Opticheskii Zhurnal):

Кореньков А.Н. Робастная регистрация облаков точек на основе метода максимального правдоподобия // Оптический журнал. 2016. Т. 83. № 7. С. 3–9.

Korenkov A.N. Robust point-cloud registration based on the maximum-likelihood method [in Russian] // Opticheskii Zhurnal. 2016. V. 83. № 7. P. 3–9.

For citation (Journal of Optical Technology):

A. Korenkov, "Robust point-cloud registration based on the maximum-likelihood method," Journal of Optical Technology. 83(7), 391-396 (2016). https://doi.org/10.1364/JOT.83.000391

Abstract:

This paper proposes a robust iterative algorithm intended for registering three-dimensional point clouds. The data to be equalized are regarded as an implementation of random quantities whose distributions are modeled by means of Gaussian mixtures. Various strategies for processing outliers in the data are considered.

Keywords:

registration, point clouds, equalization, distribution densities, maximum likelihood, outliers, robustness

Acknowledgements:

This article was prepared with the financial support of applied scientific research by the Ministry of Education and Science of the Russian Federation under contract no. 14.575.21.0055 from July 8, 2014 and the unique identifier RFMEF157514X0055 of applied scientific research.

OCIS codes: 150.0150,150.1135

References:

1. D. Chetverikov and D. Stepanov, “Robustifying the iterative closest-point algorithm,” in Proceedings of the Third Hungarian Conference on Image Processing and Pattern Recognition (KEPAF) (2002), pp. 1–9.
2. D. Chetverikov, D. Svirko, D. Stepanov, and P. Krsek, “The trimmed iterative closest-point algorithm,” in International Conference on Pattern Recognition (2002), Vol. 3, pp. 545–548.
3. A. Myronenko and X. Song, “Point-set registration: coherent point drift,” IEEE Trans. Pattern Anal. Mach. Intell. 32(12), 2262–2275 (2010).
4. A. N. Koren’kov and K. E. Monogarov, “Recording three-dimensional point clouds on the basis of the maximum-likelihood method,” Robototekh. Tekhnich. Kiber. 1(6) 46–52 (2015).
5. The Point Cloud Library (PCL) is an open project for 2D/3D image and point-cloud processing, http://pointclouds.org/documentation/tutorials/correspondence\_grouping.php#correspondence‑grouping.