RS-SLAM: RANSAC sampling for visual FastSLAM

Gim Hee Lee, Friedrich Fraundorfer, Marc Pollefeys

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

Abstract

In this paper, we present our RS-SLAM algorithm for monocular camera where the proposal distribution is derived from the 5-point RANSAC algorithm and image feature measurement uncertainties instead of using the easily violated constant velocity model. We propose to do another RANSAC sampling within all the inliers that have the best RANSAC score to check for inlier misclassifications in the original correspondences and use all the hypotheses generated from these consensus sets in the proposal distribution. This is to mitigate data association errors (inlier misclassifications) caused by the observation that the consensus set from RANSAC that yields the highest score might not, in practice, contain all the true inliers due to noise on the feature measurements. Hypotheses which are less probable will eventually be eliminated in the particle filter resampling process. We also show in this paper that our monocular approach can be easily extended for stereo camera. Experimental results validate the potential of our approach.

Original languageEnglish
Title of host publicationIROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics
Pages1655-1660
Number of pages6
DOIs
Publication statusPublished - 2011
EventInternational Conference on Intelligent Robots and Systems - San Francisco, United States
Duration: 25 Sep 201130 Sep 2011

Conference

ConferenceInternational Conference on Intelligent Robots and Systems
CountryUnited States
CitySan Francisco
Period25/09/1130/09/11

Fingerprint

Cameras
Sampling
Uncertainty

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Lee, G. H., Fraundorfer, F., & Pollefeys, M. (2011). RS-SLAM: RANSAC sampling for visual FastSLAM. In IROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics (pp. 1655-1660). [6048808] https://doi.org/10.1109/IROS.2011.6048808

RS-SLAM : RANSAC sampling for visual FastSLAM. / Lee, Gim Hee; Fraundorfer, Friedrich; Pollefeys, Marc.

IROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics. 2011. p. 1655-1660 6048808.

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

Lee, GH, Fraundorfer, F & Pollefeys, M 2011, RS-SLAM: RANSAC sampling for visual FastSLAM. in IROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics., 6048808, pp. 1655-1660, International Conference on Intelligent Robots and Systems, San Francisco, United States, 25/09/11. https://doi.org/10.1109/IROS.2011.6048808
Lee GH, Fraundorfer F, Pollefeys M. RS-SLAM: RANSAC sampling for visual FastSLAM. In IROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics. 2011. p. 1655-1660. 6048808 https://doi.org/10.1109/IROS.2011.6048808
Lee, Gim Hee ; Fraundorfer, Friedrich ; Pollefeys, Marc. / RS-SLAM : RANSAC sampling for visual FastSLAM. IROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics. 2011. pp. 1655-1660
@inproceedings{4d1b221d58f243729f295ec19f3a8ef6,
title = "RS-SLAM: RANSAC sampling for visual FastSLAM",
abstract = "In this paper, we present our RS-SLAM algorithm for monocular camera where the proposal distribution is derived from the 5-point RANSAC algorithm and image feature measurement uncertainties instead of using the easily violated constant velocity model. We propose to do another RANSAC sampling within all the inliers that have the best RANSAC score to check for inlier misclassifications in the original correspondences and use all the hypotheses generated from these consensus sets in the proposal distribution. This is to mitigate data association errors (inlier misclassifications) caused by the observation that the consensus set from RANSAC that yields the highest score might not, in practice, contain all the true inliers due to noise on the feature measurements. Hypotheses which are less probable will eventually be eliminated in the particle filter resampling process. We also show in this paper that our monocular approach can be easily extended for stereo camera. Experimental results validate the potential of our approach.",
author = "Lee, {Gim Hee} and Friedrich Fraundorfer and Marc Pollefeys",
year = "2011",
doi = "10.1109/IROS.2011.6048808",
language = "English",
isbn = "9781612844541",
pages = "1655--1660",
booktitle = "IROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics",

}

TY - GEN

T1 - RS-SLAM

T2 - RANSAC sampling for visual FastSLAM

AU - Lee, Gim Hee

AU - Fraundorfer, Friedrich

AU - Pollefeys, Marc

PY - 2011

Y1 - 2011

N2 - In this paper, we present our RS-SLAM algorithm for monocular camera where the proposal distribution is derived from the 5-point RANSAC algorithm and image feature measurement uncertainties instead of using the easily violated constant velocity model. We propose to do another RANSAC sampling within all the inliers that have the best RANSAC score to check for inlier misclassifications in the original correspondences and use all the hypotheses generated from these consensus sets in the proposal distribution. This is to mitigate data association errors (inlier misclassifications) caused by the observation that the consensus set from RANSAC that yields the highest score might not, in practice, contain all the true inliers due to noise on the feature measurements. Hypotheses which are less probable will eventually be eliminated in the particle filter resampling process. We also show in this paper that our monocular approach can be easily extended for stereo camera. Experimental results validate the potential of our approach.

AB - In this paper, we present our RS-SLAM algorithm for monocular camera where the proposal distribution is derived from the 5-point RANSAC algorithm and image feature measurement uncertainties instead of using the easily violated constant velocity model. We propose to do another RANSAC sampling within all the inliers that have the best RANSAC score to check for inlier misclassifications in the original correspondences and use all the hypotheses generated from these consensus sets in the proposal distribution. This is to mitigate data association errors (inlier misclassifications) caused by the observation that the consensus set from RANSAC that yields the highest score might not, in practice, contain all the true inliers due to noise on the feature measurements. Hypotheses which are less probable will eventually be eliminated in the particle filter resampling process. We also show in this paper that our monocular approach can be easily extended for stereo camera. Experimental results validate the potential of our approach.

UR - http://www.scopus.com/inward/record.url?scp=84455185094&partnerID=8YFLogxK

U2 - 10.1109/IROS.2011.6048808

DO - 10.1109/IROS.2011.6048808

M3 - Conference contribution

SN - 9781612844541

SP - 1655

EP - 1660

BT - IROS'11 - 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems: Celebrating 50 Years of Robotics

ER -