Generation and weighting of 3D point correspondences for improved registration of RGB-D data

K. Khoshelham, D.R. dos Santos, G. Vosselman

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

26 Citations (Scopus)
10 Downloads (Pure)


Registration of RGB-D data using visual features is often influenced by errors in the transformation of visual features to 3D space as well as the random error of individual 3D points. In a long sequence, these errors accumulate and lead to inaccurate and deformed point clouds, particularly in situations where loop closing is not feasible. We present an epipolar search method for accurate transformation of the keypoints from 2D to 3D space, and define weights for the 3D points based on the theoretical random error of depth measurements. Our results show that the epipolar search method results in more accurate 3D correspondences. We also demonstrate that weighting the 3D points improves the accuracy of sensor pose estimates along the trajectory.
Original languageEnglish
Title of host publicationWorkshop Laser Scanning 2013
Subtitle of host publication11-13 November 2013, Antalya, Turkey
EditorsM. Scaioni, R.C. Lindenbergh, S. Oude Elberink, D. Schneider, F. Pirotti
PublisherInternational Society for Photogrammetry and Remote Sensing (ISPRS)
Number of pages6
Publication statusPublished - 2013
EventISPRS Workshop Laser Scanning 2013 - Antalya, Turkey
Duration: 11 Nov 201313 Nov 2013 (Full text Open Access proceedings)

Publication series

NameISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
ISSN (Print)2194-9042


ConferenceISPRS Workshop Laser Scanning 2013
OtherISPRS Annals Volume II-5/W2
Internet address


  • Alignment
  • Indoor mapping
  • Kinect
  • Loop closing
  • Point cloud
  • RGB-D
  • SLAM


Dive into the research topics of 'Generation and weighting of 3D point correspondences for improved registration of RGB-D data'. Together they form a unique fingerprint.

Cite this