Transformer-Based Multi-Scale Feature Remote Sensing Image Classification Model

To address the challenge of poor classification accuracy due to complex backgrounds, large intra-scale variations, and high inter-scale similarity in remote sensing scene classification (RSSC), we propose a new remote sensing scene classification model called multi-scale dual-branch classification n...

Full description

Saved in:
Bibliographic Details
Main Authors: Ting Sun, Jun Li, Xiangrui Zhou, Zan Chen
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10807217/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850078575663775744
author Ting Sun
Jun Li
Xiangrui Zhou
Zan Chen
author_facet Ting Sun
Jun Li
Xiangrui Zhou
Zan Chen
author_sort Ting Sun
collection DOAJ
description To address the challenge of poor classification accuracy due to complex backgrounds, large intra-scale variations, and high inter-scale similarity in remote sensing scene classification (RSSC), we propose a new remote sensing scene classification model called multi-scale dual-branch classification network (MDBC-Net). The model is composed of a Trans-branch and CNN-branch in parallel, which can fully utilize the local attention of the CNN-branch structure and the global attention mechanism of the Trans-branch structure, thereby improving the model’s ability to focus on features of different scales. Due to the complexity of backgrounds in RSSC, we require features at different scales to obtain richer scene information. Thus we design a down-sampling module in the model to obtain multi-scale features. Finally, we adopt the polynomial form of cross entropy for the trained loss function to improve the generalization performance and robustness of the model. Experiments have shown that the model achieves advanced performance on three datasets: NWPU-RESISC45, AID, and UC Served.
format Article
id doaj-art-42b37eb41f324538bd0d5a645c97e24c
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-42b37eb41f324538bd0d5a645c97e24c2025-08-20T02:45:31ZengIEEEIEEE Access2169-35362025-01-0113340953410410.1109/ACCESS.2024.352025310807217Transformer-Based Multi-Scale Feature Remote Sensing Image Classification ModelTing Sun0https://orcid.org/0009-0009-2956-1577Jun Li1https://orcid.org/0009-0004-1683-5130Xiangrui Zhou2Zan Chen3https://orcid.org/0000-0003-4252-4761School of Culture and Tourism, Zhejiang International Studies University, Hangzhou, ChinaCollege of Information Engineering, Zhejiang University of Technology, Hangzhou, ChinaCollege of Information Engineering, Zhejiang University of Technology, Hangzhou, ChinaCollege of Information Engineering, Zhejiang University of Technology, Hangzhou, ChinaTo address the challenge of poor classification accuracy due to complex backgrounds, large intra-scale variations, and high inter-scale similarity in remote sensing scene classification (RSSC), we propose a new remote sensing scene classification model called multi-scale dual-branch classification network (MDBC-Net). The model is composed of a Trans-branch and CNN-branch in parallel, which can fully utilize the local attention of the CNN-branch structure and the global attention mechanism of the Trans-branch structure, thereby improving the model’s ability to focus on features of different scales. Due to the complexity of backgrounds in RSSC, we require features at different scales to obtain richer scene information. Thus we design a down-sampling module in the model to obtain multi-scale features. Finally, we adopt the polynomial form of cross entropy for the trained loss function to improve the generalization performance and robustness of the model. Experiments have shown that the model achieves advanced performance on three datasets: NWPU-RESISC45, AID, and UC Served.https://ieeexplore.ieee.org/document/10807217/Remote sensingscene classificationmulti-scale featureslocal attentionglobal attentionpolynomial form
spellingShingle Ting Sun
Jun Li
Xiangrui Zhou
Zan Chen
Transformer-Based Multi-Scale Feature Remote Sensing Image Classification Model
IEEE Access
Remote sensing
scene classification
multi-scale features
local attention
global attention
polynomial form
title Transformer-Based Multi-Scale Feature Remote Sensing Image Classification Model
title_full Transformer-Based Multi-Scale Feature Remote Sensing Image Classification Model
title_fullStr Transformer-Based Multi-Scale Feature Remote Sensing Image Classification Model
title_full_unstemmed Transformer-Based Multi-Scale Feature Remote Sensing Image Classification Model
title_short Transformer-Based Multi-Scale Feature Remote Sensing Image Classification Model
title_sort transformer based multi scale feature remote sensing image classification model
topic Remote sensing
scene classification
multi-scale features
local attention
global attention
polynomial form
url https://ieeexplore.ieee.org/document/10807217/
work_keys_str_mv AT tingsun transformerbasedmultiscalefeatureremotesensingimageclassificationmodel
AT junli transformerbasedmultiscalefeatureremotesensingimageclassificationmodel
AT xiangruizhou transformerbasedmultiscalefeatureremotesensingimageclassificationmodel
AT zanchen transformerbasedmultiscalefeatureremotesensingimageclassificationmodel