Histogram Based Visual Place Recognition for Improving SLAM Performance

In SLAM systems, loop closures play crucial role to decrease the accumulating drift of the estimated trajectory for consistent mapping. Place recognition is so important for determining loop closure candidates. This paper presents a method for visual place recognition using image histograms in conju...

Full description

Saved in:
Bibliographic Details
Published in2016 International Conference on Autonomous Robot Systems and Competitions (ICARSC) pp. 174 - 180
Main Authors Guclu, Oguzhan, Can, Ahmet Burak
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.05.2016
Subjects
Cameras
histogram
Histograms
keypoint matching
loop closure
Optimization
Simultaneous localization and mapping
SLAM
Trajectory
Visualization
Online AccessGet full text

Cover

More Information
Summary:In SLAM systems, loop closures play crucial role to decrease the accumulating drift of the estimated trajectory for consistent mapping. Place recognition is so important for determining loop closure candidates. This paper presents a method for visual place recognition using image histograms in conjunction with keypoint correspondences to select loop closure candidates. The method compares color and grayscale histograms of RGB images and then applies keypoint matching to the images having similar histograms. The proposed method is integrated into a state-of-the-art RGB-D SLAM system and tested on a popular dataset. The quantitative results show that the proposed method improves accuracy significantly.
DOI:10.1109/ICARSC.2016.42