Browse > Article
http://dx.doi.org/10.6109/jkiice.2021.25.5.652

Effective criterion for evaluating registration accuracy  

Lim, Sukhyun (Innovation Center, 3D Systems Korea)
Publication Information
Journal of the Korea Institute of Information and Communication Engineering / v.25, no.5, 2021 , pp. 652-658 More about this Journal
Abstract
When acquiring a point cloud using a 3D scanner, a registration process of making the acquired data based on each local coordinate into one data with a unified world coordinate system is required. Its process is difficult to obtain a satisfactory result with only one execution, and it is repeated several times to increase the registration precision. The criterion for determining the registration accuracy is an important factor. The previous methods for determining the accuracy of registration have a limitation in that the judgment may be ambiguous in some cases, and different results may be produced each time depending on the characteristics of the point cloud. Therefore, to calculate the accuracy of registration more precisely, I propose a method using the average distance value of the point group for the entire points rather than the corresponding points used in the registration. When this method is used, it is possible to determine the registration accuracy more reliably than the conventional methods.
Keywords
Registration; ICP; RMSE; MAE;
Citations & Related Records
연도 인용수 순위
  • Reference
1 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.   DOI
2 P. J. Besl and M. D. McKay, "Method for registration of 3-D shapes," International Society for Optics and Photonics, vol. 1611, pp. 586-607, 1992.
3 F. Deng, "Registration between multiple laser scanner data sets," in Laser Scanning, Theory and Applications, pp. 449-472, 2011.
4 K. Y. Kown, "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.   DOI
5 J. Zhang, Y. Yao, and B. Deng, "Fast and Robust Iterative Closest Point," IEEE Transactions on Pattern Analysis and Machine Intelligence, preprint, 2021.
6 B. Becerik-Gerber, F. Jazizadeh, G. Kavulya, and G. Calis, "Assessment of target types and layouts in 3D laser scanning for registration accuracy," Automation in Construction, vol. 20, no. 5, pp. 649-658, 2011.   DOI
7 T. H. Cho, "Efficient CUDA Implementation of Multiple Planes Fitting Using RANSAC," Journal of the Korea Institute of Information and Communication Engineering, vol. 23, no. 4, pp. 388-393, 2019.   DOI
8 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.   DOI
9 D. H. Yun, S. I. Choi, S. H. Kim, and K. H. Ko, "Registration of multiview point clouds for application to ship fabrication," Graphical Models, vol. 90, pp. 1-12, 2017.   DOI
10 S. Rusinkiewicz, "A Symmetric Objective Function for ICP," ACM Transactions on Graphics, vol. 38, no. 4, pp. 1-7, 2019.   DOI
11 N. Mellado, D. Aiger, and N. J. Mitra, "Fast global pointcloud registration via smart indexing," Computer Graphics Forum, vol. 33, no. 5, pp. 205-215, 2014.   DOI
12 Y. He, B. Liang, J. Yang, S. Li, and J. He, "An iterative closest points algorithm for registration of 3D laser scanner point clouds with geometric features," Sensors, vol. 17, no. 8, 2017.
13 Stanford University 3D Scan Repository [Internet]. Available: http://graphics.stanford.edu/data/.