Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.11851/3891
Title: A New Navigation System for Unmanned Aerial Vehicles in Global Positioning System-Denied Environments Based On Image Registration with Mutual Information and Deep Learning
Authors: Şahin, Çağla
Yetik, İmam Şamil
Keywords: Simultaneous Localization and Mapping 
 Ostdeutscher Rundfunk Brandenburg 
 Pose Estimation
Issue Date: Jan-2020
Publisher: Institute of Navigation
Source: Şahin, Ç. and Yetik, İ. Ş. (2020) A New Navigation System for Unmanned Aerial Vehicles in Global Positioning System-Denied Environments Based On Image Registration with Mutual Information and Deep Learning. Institute of Navigation.
Abstract: In this paper, we develop an alternative navigation system for Unmanned Aerial Vehicle (UAV) in Global Positioning Systems (GPS)-denied environment. We use two image inputs, one is acquired with an on-board camera placed on the UAV (which is the large-area image) and the other is from satellite images (which is small known image) with GPS information. We use a convolutional neural network (CNN) architecture based on Oxford's Visual Geometry Group network (VGG-16) and utilize normalized variant mutual information between these two images to obtain position of the UAV. Satellite images are labelled and given to the UAV. When GPS information is lost, our algorithm starts to function and images from UAV camera are searched whether satellite image is seen by cameras on UAV image or not. If the UAV is in that area, our algorithm finds the GPS information from satellite image data. © 2020 ION 2020 International Technical Meeting Proceedings. All rights reserved.
URI: https://hdl.handle.net/20.500.11851/3891
https://www.ion.org/publications/abstract.cfm?articleID=17203
ISBN: 0936406240
978-093640624-4
Appears in Collections:Elektrik ve Elektronik Mühendisliği Bölümü / Department of Electrical & Electronics Engineering
Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection
WoS İndeksli Yayınlar Koleksiyonu / WoS Indexed Publications Collection

Show full item record

CORE Recommender

Page view(s)

30
checked on Dec 26, 2022

Google ScholarTM

Check

Altmetric


Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.