Absolute scale estimation of ORB-SLAM algorithm based on laser ranging

Aiming at the problem that the visual algorithm ORB-SLAM can locate accurately only in the set scale or known environment to initial which caused by that it can't directly restore absolute scale, an absolute scale estimation of ORB-SLAM algorithm based on laser ranging is proposed in this paper...

Full description

Saved in:
Bibliographic Details
Main Authors: Lv, Qiang, Ma, Jianye, Wang, Guosheng, Lin, Huican
Format: Conference Proceeding
Language:English
Subjects:
Absolute Scale
Accurate Localization
Algorithms
Calibration
Cameras
Conferences
Feature extraction
Indoor environments
Laser modes
Laser range finders
Laser Ranging
Lasers
Localization
Measurement by laser beam
ORB-SLAM
Position (location)
Robot sensing systems
Unknown environments
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Aiming at the problem that the visual algorithm ORB-SLAM can locate accurately only in the set scale or known environment to initial which caused by that it can't directly restore absolute scale, an absolute scale estimation of ORB-SLAM algorithm based on laser ranging is proposed in this paper. Due to there is a certain relationship between scale and distance, the absolute scale is estimated by the weighted average of distances that are measured within a given range by laser range finder, the ORB-SLAM algorithm can achieve accurate localization at any distance to initial in unknown environment. Experiments show that the system can achieve accurate localization in the indoor planar environment and indoor non-planar environment, and the planar environment is more accurate.
ISSN:2161-2927
1934-1768
DOI:10.1109/ChiCC.2016.7554983