Citation: Zhang, W.; Qu, J.; Wang, W.;
Hu, J.; Li, J. Geo-Location Method for
Images of Damaged Roads.
Electronics 2022, 11, 2530. https://
doi.org/10.3390/electronics11162530
Academic Editor: Byung Cheol Song
Received: 21 July 2022
Accepted: 10 August 2022
Published: 12 August 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Geo-Location Method for Images of Damaged Roads
Wenbo Zhang , Jue Qu, Wei Wang *, Jun Hu and Jie Li
Air and Missile Defense College, Air Force Engineering University, Xi’an 710051, China
* Correspondence: wangwei_afeuv@163.com
Abstract:
Due to the large difference between normal conditions and damaged road images, geo-
location in damaged areas often fails due to occlusion or damage to buildings and iconic signage in
the image. In order to study the influence of post-war building and landmark damage conditions
on the geolocation results of localization algorithms, and to improve the geolocation effect of such
algorithms under damaged conditions, this paper used informative reference images and key point
selection. Aiming at the negative effects of occlusion and landmark building damage in the retrieval
process, a retrieval method called reliability- and repeatability-based deep learning feature points
is proposed. In order to verify the effectiveness of the above algorithm, this paper constructed a
data set consisting of urban, rural and technological parks and other road segments as a training
set to generate a database. It consists of 11,896 reference images. Considering the cost of damaged
landmarks, an artificially generated method is used to construct images of damaged landmarks with
different damage ratios as a test set. Experiments show that the database optimization method can
effectively compress the storage capacity of the feature index and can also speed up the positioning
speed without affecting the accuracy rate. The proposed image retrieval method optimizes feature
points and feature indices to make them reliable against damaged terrain and images. The improved
algorithm improved the accuracy of geo-location for damaged roads, and the method based on deep
learning has a higher effect on the geo-location of damaged roads than the traditional algorithm.
Furthermore, we fully demonstrated the effectiveness of our proposed method by constructing a
multi-segment road image dataset.
Keywords:
visual place recognition; damaged road images; database optimization; image
breakage method
1. Introduction
The Global Navigation Satellite System (GNSS) has been a relatively mature posi-
tioning technology, and it has been widely used in many fields. In the actual use of
the GNSS positioning method, the ground signal strength is weak, and the civil coding
structure is open, which makes the GNSS signal vulnerable to interference under complex
electromagnetic interference and is prone to positioning failure when it is interfered by
malicious deception. Although nowadays, smart devices have the function of obtaining
geographic tags, when the GNSS signal is interfered and still needs to be positioned, users
cannot rely on GNSS to obtain positioning information, and they are located in cities,
post-disaster damaged areas, and geodetic control points. Being damaged and unable to
be used normally, the inertial navigation equipment cannot be effectively calibrated, so
it is difficult to use the inertial navigation equipment for precise positioning. In recent
years, visual place recognition has received great attention in the field of machine vision,
which can be used to solve the problem of location information localization. If relatively
accurate geo-location information is added to these images, they can be of great benefit in
areas such as outdoor localization [
1
], pedestrian detection [
2
], autonomous driving [
3
], etc.
In addition, pictures with geolocation information can also help environment perception
technology for robots [
4
] and urban construction. Therefore, it is a problem that needs to be
researched to identify the visual position of damaged road images under the condition of
interference with GNSS signals, and to perform geolocation at the same time.
Electronics 2022, 11, 2530. https://doi.org/10.3390/electronics11162530 https://www.mdpi.com/journal/electronics
