ORB-SHOT SLAM

Trajectory correction by 3D loop closing based on bag-of-visual-words (BoVM) model for RGBb-D visual SLAM

Zheng Chai, Takafumi Matsumaru

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This paper proposes the ORB-SHOT SLAM or OS-SLAM, which is a novel method of 3D loop closing for trajectory correction of RGB-D visual SLAM. We obtain point clouds from RGB-D sensors such as Kinect or Xtion, and we use 3D SHOT descriptors to describe the ORB corners. Then, we train an offline 3D vocabulary that contains more than 600,000 words by using two million 3D descriptors based on a large number of images from a public dataset provided by TUM. We convert new images to bag-of-visual-words (BoVW) vectors and push these vectors into an incremental database. We query the database for new images to detect the corresponding 3D loop candidates, and compute similarity scores between the new image and each corresponding 3D loop candidate. After detecting 2D loop closures using ORB-SLAM2 system, we accept those loop closures that are also included in the 3D loop candidates, and we assign them corresponding weights according to the scores stored previously. In the final graph-based optimization, we create edges with different weights for loop closures and correct the trajectory by solving a nonlinear least-squares optimization problem. We compare our results with several state-of-the-art systems such as ORB-SLAM2 and RGB-D SLAM by using the TUM public RGB-D dataset. We find that accurate loop closures and suitable weights reduce the error on trajectory estimation more effectively than other systems. The performance of ORB-SHOT SLAM is demonstrated by 3D reconstruction application.

Original languageEnglish
Pages (from-to)365-380
Number of pages16
JournalJournal of Robotics and Mechatronics
Volume29
Issue number2
DOIs
Publication statusPublished - 2017 Apr 1

Fingerprint

Trajectories
Sensors

Keywords

  • 3D vocabulary
  • Bag-of-visual-words (BoVW)
  • Loop closing
  • RGB-D SLAM
  • Trajectory correction

ASJC Scopus subject areas

  • Computer Science(all)
  • Electrical and Electronic Engineering

Cite this

@article{33031b30293645d5b273399821c34aef,
title = "ORB-SHOT SLAM: Trajectory correction by 3D loop closing based on bag-of-visual-words (BoVM) model for RGBb-D visual SLAM",
abstract = "This paper proposes the ORB-SHOT SLAM or OS-SLAM, which is a novel method of 3D loop closing for trajectory correction of RGB-D visual SLAM. We obtain point clouds from RGB-D sensors such as Kinect or Xtion, and we use 3D SHOT descriptors to describe the ORB corners. Then, we train an offline 3D vocabulary that contains more than 600,000 words by using two million 3D descriptors based on a large number of images from a public dataset provided by TUM. We convert new images to bag-of-visual-words (BoVW) vectors and push these vectors into an incremental database. We query the database for new images to detect the corresponding 3D loop candidates, and compute similarity scores between the new image and each corresponding 3D loop candidate. After detecting 2D loop closures using ORB-SLAM2 system, we accept those loop closures that are also included in the 3D loop candidates, and we assign them corresponding weights according to the scores stored previously. In the final graph-based optimization, we create edges with different weights for loop closures and correct the trajectory by solving a nonlinear least-squares optimization problem. We compare our results with several state-of-the-art systems such as ORB-SLAM2 and RGB-D SLAM by using the TUM public RGB-D dataset. We find that accurate loop closures and suitable weights reduce the error on trajectory estimation more effectively than other systems. The performance of ORB-SHOT SLAM is demonstrated by 3D reconstruction application.",
keywords = "3D vocabulary, Bag-of-visual-words (BoVW), Loop closing, RGB-D SLAM, Trajectory correction",
author = "Zheng Chai and Takafumi Matsumaru",
year = "2017",
month = "4",
day = "1",
doi = "10.20965/jrm.2017.p0365",
language = "English",
volume = "29",
pages = "365--380",
journal = "Journal of Robotics and Mechatronics",
issn = "0915-3942",
publisher = "Fuji Technology Press",
number = "2",

}

TY - JOUR

T1 - ORB-SHOT SLAM

T2 - Trajectory correction by 3D loop closing based on bag-of-visual-words (BoVM) model for RGBb-D visual SLAM

AU - Chai, Zheng

AU - Matsumaru, Takafumi

PY - 2017/4/1

Y1 - 2017/4/1

N2 - This paper proposes the ORB-SHOT SLAM or OS-SLAM, which is a novel method of 3D loop closing for trajectory correction of RGB-D visual SLAM. We obtain point clouds from RGB-D sensors such as Kinect or Xtion, and we use 3D SHOT descriptors to describe the ORB corners. Then, we train an offline 3D vocabulary that contains more than 600,000 words by using two million 3D descriptors based on a large number of images from a public dataset provided by TUM. We convert new images to bag-of-visual-words (BoVW) vectors and push these vectors into an incremental database. We query the database for new images to detect the corresponding 3D loop candidates, and compute similarity scores between the new image and each corresponding 3D loop candidate. After detecting 2D loop closures using ORB-SLAM2 system, we accept those loop closures that are also included in the 3D loop candidates, and we assign them corresponding weights according to the scores stored previously. In the final graph-based optimization, we create edges with different weights for loop closures and correct the trajectory by solving a nonlinear least-squares optimization problem. We compare our results with several state-of-the-art systems such as ORB-SLAM2 and RGB-D SLAM by using the TUM public RGB-D dataset. We find that accurate loop closures and suitable weights reduce the error on trajectory estimation more effectively than other systems. The performance of ORB-SHOT SLAM is demonstrated by 3D reconstruction application.

AB - This paper proposes the ORB-SHOT SLAM or OS-SLAM, which is a novel method of 3D loop closing for trajectory correction of RGB-D visual SLAM. We obtain point clouds from RGB-D sensors such as Kinect or Xtion, and we use 3D SHOT descriptors to describe the ORB corners. Then, we train an offline 3D vocabulary that contains more than 600,000 words by using two million 3D descriptors based on a large number of images from a public dataset provided by TUM. We convert new images to bag-of-visual-words (BoVW) vectors and push these vectors into an incremental database. We query the database for new images to detect the corresponding 3D loop candidates, and compute similarity scores between the new image and each corresponding 3D loop candidate. After detecting 2D loop closures using ORB-SLAM2 system, we accept those loop closures that are also included in the 3D loop candidates, and we assign them corresponding weights according to the scores stored previously. In the final graph-based optimization, we create edges with different weights for loop closures and correct the trajectory by solving a nonlinear least-squares optimization problem. We compare our results with several state-of-the-art systems such as ORB-SLAM2 and RGB-D SLAM by using the TUM public RGB-D dataset. We find that accurate loop closures and suitable weights reduce the error on trajectory estimation more effectively than other systems. The performance of ORB-SHOT SLAM is demonstrated by 3D reconstruction application.

KW - 3D vocabulary

KW - Bag-of-visual-words (BoVW)

KW - Loop closing

KW - RGB-D SLAM

KW - Trajectory correction

UR - http://www.scopus.com/inward/record.url?scp=85018708723&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85018708723&partnerID=8YFLogxK

U2 - 10.20965/jrm.2017.p0365

DO - 10.20965/jrm.2017.p0365

M3 - Article

VL - 29

SP - 365

EP - 380

JO - Journal of Robotics and Mechatronics

JF - Journal of Robotics and Mechatronics

SN - 0915-3942

IS - 2

ER -