Feature-based monocular real-time localization for UAVs in indoor environment

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

Abstract

A real-time localization method based on monocular is proposed for unmanned aerial vehicles (UAVs) navigation in the indoor environment. ORB features are used to speed up the feature detection and feature matching is made more accurate by applying random sampling consensus strategies. We take advantage of g2o (General Graphic Optimization) to realize the bundle adjustment and thus improve the precision of motion estimation. To reduce the drift of estimated trajectory to a certain extent, a map based algorithm is adopted instead of the traditional pairwise visual odometry. Finally, the effectiveness of the algorithm is verified by the dataset, and several experiments have been performed in indoor environment to evaluate the performance of the algorithm.

Original languageEnglish
Title of host publicationProceedings of 2017 Chinese Intelligent Automation Conference
EditorsZhidong Deng
PublisherSpringer Verlag
Pages357-366
Number of pages10
ISBN (Print)9789811064449
DOIs
StatePublished - 2018
EventChinese Intelligent Automation Conference, CIAC 2017 - Tianjin, China
Duration: 2 Jun 20174 Jun 2017

Publication series

NameLecture Notes in Electrical Engineering
Volume458
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Conference

ConferenceChinese Intelligent Automation Conference, CIAC 2017
Country/TerritoryChina
CityTianjin
Period2/06/174/06/17

Keywords

  • Indoor environment
  • Monocular vision navigation
  • Real-time localization
  • UAV

Fingerprint

Dive into the research topics of 'Feature-based monocular real-time localization for UAVs in indoor environment'. Together they form a unique fingerprint.

Cite this