• Sorted by Date • Classified by Publication Type • Classified by Research Category •
Ecenur Oğuz, Yalım Doğan, Uğur Güdükbay, Oya Karaşan, and Mustafa Pınar. Point Cloud Registration with Quantile Assignment. Machine Vision and Applications, 35(3):Article No. 38, 17 pages, May 2024.
Point cloud registration is a fundamental problem in computer vision. The problem encompasses critical tasks such as feature estimation, correspondence matching, and transformation estimation. The point cloud registration problem can be cast as aquantile matching problem. We refined the quantile assignment algorithm by integrating prevalent feature descriptors and transformation estimation methods to enhance the correspondence between the source and target point clouds. We evaluated theperformances of these descriptors and methods with our approach through controlled experiments on a dataset we constructed using well-known 3D models. This systematic investigation led us to identify the most suitable methods for complementingour approach. Subsequently, we devised a new end-to-end, coarse-to-fine pairwise point cloud registration framework. Finally, we tested our framework on indoor and outdoor benchmark datasets and compared our results with state-of-the-art point cloud registration methods.
@Article{OguzDGKP24, author = {Ecenur O{\^g}uz and Yal{\i}m Do{\^g}an and U{\^g}ur G{\"u}d{\"u}kbay and Oya Kara{\c s}an and Mustafa P{\i}nar}, title = {Point Cloud Registration with Quantile Assignment}, journal = {Machine Vision and Applications}, volume = {35}, number = {3}, year = {2024}, month = {May}, pages = {Article No. 38, 17 pages}, keywords = {Point cloud registration, Fast point feature histograms (FPFH) descriptor, Quantile assignment, Iterative closest-point algorithm, Bipartite graph matching, Hungarian algorithm, Hopcroft-Karp algorithm}, abstract = {Point cloud registration is a fundamental problem in computer vision. The problem encompasses critical tasks such as feature estimation, correspondence matching, and transformation estimation. The point cloud registration problem can be cast as aquantile matching problem. We refined the quantile assignment algorithm by integrating prevalent feature descriptors and transformation estimation methods to enhance the correspondence between the source and target point clouds. We evaluated theperformances of these descriptors and methods with our approach through controlled experiments on a dataset we constructed using well-known 3D models. This systematic investigation led us to identify the most suitable methods for complementingour approach. Subsequently, we devised a new end-to-end, coarse-to-fine pairwise point cloud registration framework. Finally, we tested our framework on indoor and outdoor benchmark datasets and compared our results with state-of-the-art point cloud registration methods.}, ee = {https://doi.org/10.1007/s00138-024-01517-3}, bib2html_dl_pdf = "http://www.cs.bilkent.edu.tr/~gudukbay/publications/papers/journal_articles/Oguz_Et_Al_MVA_2024.pdf", bib2html_pubtype = {Refereed Journal Articles}, bib2html_rescat = {Computer Vision}, bibsource = {DBLP, http://dblp.uni-trier.de} }
Generated by bib2html.pl (written by Patrick Riley ) on Sun Apr 21, 2024 11:32:41