Aerial Images Segmentation with Graph Neural Network

The development of remote sensing platforms and sensors, as well as the improvement of remote data processing tools and methods, create new opportunities for automatic updating of maps. Currently, aerial photographs serve as the main source for automatic map updates due to their accessibility and si...

Full description

Saved in:
Bibliographic Details
Main Authors: A. V. Emelyanov, V. A. Knyaz, V. V. Kniaz, S. Yu. Zheltov
Format: Article
Language:English
Published: Copernicus Publications 2024-12-01
Series:ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Online Access:https://isprs-annals.copernicus.org/articles/X-2-W1-2024/1/2024/isprs-annals-X-2-W1-2024-1-2024.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846119696859922432
author A. V. Emelyanov
A. V. Emelyanov
V. A. Knyaz
V. A. Knyaz
V. V. Kniaz
V. V. Kniaz
S. Yu. Zheltov
author_facet A. V. Emelyanov
A. V. Emelyanov
V. A. Knyaz
V. A. Knyaz
V. V. Kniaz
V. V. Kniaz
S. Yu. Zheltov
author_sort A. V. Emelyanov
collection DOAJ
description The development of remote sensing platforms and sensors, as well as the improvement of remote data processing tools and methods, create new opportunities for automatic updating of maps. Currently, aerial photographs serve as the main source for automatic map updates due to their accessibility and significant informational value. One of the core elements for image to maps transition is accurate image segmentation. Nowadays, machine learning methods demonstrate the best results in task of image segmentation. At its core, maps represent information about a certain area in a vector form, that not only contains visual information about area, but also reflects some relations between objects in the map. This quality makes a map more convenient for human perception than an aerial photograph (raster image). This study addresses the problem of accurate aerial image segmentation with taking the advantages of using graph neural network as the more adequate model of map structure. We use graph neural network for retrieving semantic and vector information about a captured area from its aerial image. The developed framework at first phase utilizes visual transformer for retrieving deep features from the input aerial image. The graph neural network then performs clustering of the extracted deep features to obtain semantic segmentation of the image. To train and evaluate the developed framework, a special dataset is collected and annotated. It contains more than 10k aerial photographs representing various types of objects taken in different years and seasons. The evaluation results on the created dataset proved the state-of-the-art performance of the developed framework.
format Article
id doaj-art-5fb373391d834e1f95d307d88f6bf37a
institution Kabale University
issn 2194-9042
2194-9050
language English
publishDate 2024-12-01
publisher Copernicus Publications
record_format Article
series ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
spelling doaj-art-5fb373391d834e1f95d307d88f6bf37a2024-12-16T22:28:19ZengCopernicus PublicationsISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences2194-90422194-90502024-12-01X-2-W1-20241810.5194/isprs-annals-X-2-W1-2024-1-2024Aerial Images Segmentation with Graph Neural NetworkA. V. Emelyanov0A. V. Emelyanov1V. A. Knyaz2V. A. Knyaz3V. V. Kniaz4V. V. Kniaz5S. Yu. Zheltov6Moscow Institute of Physics and Technology (MIPT), Moscow, RussiaState Research Institute of Aviation Systems (GosNIIAS), Moscow, RussiaMoscow Institute of Physics and Technology (MIPT), Moscow, RussiaState Research Institute of Aviation Systems (GosNIIAS), Moscow, RussiaMoscow Institute of Physics and Technology (MIPT), Moscow, RussiaState Research Institute of Aviation Systems (GosNIIAS), Moscow, RussiaState Research Institute of Aviation Systems (GosNIIAS), Moscow, RussiaThe development of remote sensing platforms and sensors, as well as the improvement of remote data processing tools and methods, create new opportunities for automatic updating of maps. Currently, aerial photographs serve as the main source for automatic map updates due to their accessibility and significant informational value. One of the core elements for image to maps transition is accurate image segmentation. Nowadays, machine learning methods demonstrate the best results in task of image segmentation. At its core, maps represent information about a certain area in a vector form, that not only contains visual information about area, but also reflects some relations between objects in the map. This quality makes a map more convenient for human perception than an aerial photograph (raster image). This study addresses the problem of accurate aerial image segmentation with taking the advantages of using graph neural network as the more adequate model of map structure. We use graph neural network for retrieving semantic and vector information about a captured area from its aerial image. The developed framework at first phase utilizes visual transformer for retrieving deep features from the input aerial image. The graph neural network then performs clustering of the extracted deep features to obtain semantic segmentation of the image. To train and evaluate the developed framework, a special dataset is collected and annotated. It contains more than 10k aerial photographs representing various types of objects taken in different years and seasons. The evaluation results on the created dataset proved the state-of-the-art performance of the developed framework.https://isprs-annals.copernicus.org/articles/X-2-W1-2024/1/2024/isprs-annals-X-2-W1-2024-1-2024.pdf
spellingShingle A. V. Emelyanov
A. V. Emelyanov
V. A. Knyaz
V. A. Knyaz
V. V. Kniaz
V. V. Kniaz
S. Yu. Zheltov
Aerial Images Segmentation with Graph Neural Network
ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
title Aerial Images Segmentation with Graph Neural Network
title_full Aerial Images Segmentation with Graph Neural Network
title_fullStr Aerial Images Segmentation with Graph Neural Network
title_full_unstemmed Aerial Images Segmentation with Graph Neural Network
title_short Aerial Images Segmentation with Graph Neural Network
title_sort aerial images segmentation with graph neural network
url https://isprs-annals.copernicus.org/articles/X-2-W1-2024/1/2024/isprs-annals-X-2-W1-2024-1-2024.pdf
work_keys_str_mv AT avemelyanov aerialimagessegmentationwithgraphneuralnetwork
AT avemelyanov aerialimagessegmentationwithgraphneuralnetwork
AT vaknyaz aerialimagessegmentationwithgraphneuralnetwork
AT vaknyaz aerialimagessegmentationwithgraphneuralnetwork
AT vvkniaz aerialimagessegmentationwithgraphneuralnetwork
AT vvkniaz aerialimagessegmentationwithgraphneuralnetwork
AT syuzheltov aerialimagessegmentationwithgraphneuralnetwork