Looking around in the neighbourhood: Location estimation of outdoor urban images
Abstract Visual geolocalisation has remained as a challenge in the research community: Given a query image, and a geo‐tagged reference database, the goal is to derive a location estimate for the query image. We propose an approach to tackling the geolocalisation problem in a four‐step manner. Essent...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2021-08-01
|
Series: | IET Image Processing |
Online Access: | https://doi.org/10.1049/ipr2.12190 |
id |
doaj-8302dffe00ad4b59ba24448f25df4e02 |
---|---|
record_format |
Article |
spelling |
doaj-8302dffe00ad4b59ba24448f25df4e022021-07-22T05:40:40ZengWileyIET Image Processing1751-96591751-96672021-08-0115102227223910.1049/ipr2.12190Looking around in the neighbourhood: Location estimation of outdoor urban imagesJie Huang0Haozhi Huang1Faculty of Information Technology Macau University of Science and Technology Avenida Wai Long Macau ChinaSchool of Computer Science Fudan University Shanghai ChinaAbstract Visual geolocalisation has remained as a challenge in the research community: Given a query image, and a geo‐tagged reference database, the goal is to derive a location estimate for the query image. We propose an approach to tackling the geolocalisation problem in a four‐step manner. Essentially, our approach focuses on re‐ranking the candidate images after image retrieval, by considering the visual similarity of the candidate and its neighbouring images, to the query image. By introducing the neighbouring images, the visual information of a candidate location has been enriched. The evaluation has been conducted on three street view datasets, where our approach outperforms three baseline approaches, in terms of location estimation accuracy on two datasets. We provide discussions related to, firstly, whether using deep features for image retrieval helps improve location estimation accuracy, and the effectiveness of geographical neighbourhoods; secondly, using different deep architectures for feature extraction, and its impact on estimation accuracy; thirdly, investigating if our approach consistently outperforms the classic 1‐NN approach, on two datasets with significant difference in visual elements.https://doi.org/10.1049/ipr2.12190 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Jie Huang Haozhi Huang |
spellingShingle |
Jie Huang Haozhi Huang Looking around in the neighbourhood: Location estimation of outdoor urban images IET Image Processing |
author_facet |
Jie Huang Haozhi Huang |
author_sort |
Jie Huang |
title |
Looking around in the neighbourhood: Location estimation of outdoor urban images |
title_short |
Looking around in the neighbourhood: Location estimation of outdoor urban images |
title_full |
Looking around in the neighbourhood: Location estimation of outdoor urban images |
title_fullStr |
Looking around in the neighbourhood: Location estimation of outdoor urban images |
title_full_unstemmed |
Looking around in the neighbourhood: Location estimation of outdoor urban images |
title_sort |
looking around in the neighbourhood: location estimation of outdoor urban images |
publisher |
Wiley |
series |
IET Image Processing |
issn |
1751-9659 1751-9667 |
publishDate |
2021-08-01 |
description |
Abstract Visual geolocalisation has remained as a challenge in the research community: Given a query image, and a geo‐tagged reference database, the goal is to derive a location estimate for the query image. We propose an approach to tackling the geolocalisation problem in a four‐step manner. Essentially, our approach focuses on re‐ranking the candidate images after image retrieval, by considering the visual similarity of the candidate and its neighbouring images, to the query image. By introducing the neighbouring images, the visual information of a candidate location has been enriched. The evaluation has been conducted on three street view datasets, where our approach outperforms three baseline approaches, in terms of location estimation accuracy on two datasets. We provide discussions related to, firstly, whether using deep features for image retrieval helps improve location estimation accuracy, and the effectiveness of geographical neighbourhoods; secondly, using different deep architectures for feature extraction, and its impact on estimation accuracy; thirdly, investigating if our approach consistently outperforms the classic 1‐NN approach, on two datasets with significant difference in visual elements. |
url |
https://doi.org/10.1049/ipr2.12190 |
work_keys_str_mv |
AT jiehuang lookingaroundintheneighbourhoodlocationestimationofoutdoorurbanimages AT haozhihuang lookingaroundintheneighbourhoodlocationestimationofoutdoorurbanimages |
_version_ |
1721292091457273856 |