Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover Classification

Land cover classification has the goal to attribute each pixel of high-resolution remoste sensing image with planimetric category labels (such as vegetation, building, and water). In recent years, many serial deep-learning architectures (features are delivered through a single path, such as in <i...

Full description

Saved in:
Bibliographic Details
Main Authors: Zongqian Zhan, Zirou Xiong, Xin Huang, Chun Yang, Yi Liu, Xin Wang
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10356620/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850040774274580480
author Zongqian Zhan
Zirou Xiong
Xin Huang
Chun Yang
Yi Liu
Xin Wang
author_facet Zongqian Zhan
Zirou Xiong
Xin Huang
Chun Yang
Yi Liu
Xin Wang
author_sort Zongqian Zhan
collection DOAJ
description Land cover classification has the goal to attribute each pixel of high-resolution remoste sensing image with planimetric category labels (such as vegetation, building, and water). In recent years, many serial deep-learning architectures (features are delivered through a single path, such as in <italic>ResNet</italic>, <italic>MobileNet</italic>, and <italic>Segformer</italic>) based on convolutional neural networks and attention mechanisms have been widely explored in land cover classification. However, high-resolution remote sensing images typically have abundant textual details, variable scales in objects, large intraclass variance, and similar interclass correlation, which bring challenges to land cover classification. In this work, we present two pluggable modules to further boost serial learning architecture: first, to cope with ambiguous boundaries caused by lost details and fragmented segmentation stemmed from scale variances, a combination of spatial attention and channel attention is proposed for multiscale feature reconstruction (MSFR); second, to mitigate the classification error caused by intraclass variance and interclass correlation, we explore an interclass attention weighting (ICAW) module, which builds feature vectors for each category, and applies a multihead attention model to capture self-attention dependence among different categories. The experimental results demonstrate that the proposed modules are feasible to the existing serial learning architectures and can improve overall accuracy (OA) by 5.64&#x0025; on the ISPRS Vaihingen two-dimensional dataset (using <italic>ResNet50</italic> as a backbone); in particular, the OA values are 80.68&#x0025; and 86.32&#x0025; before and after using the proposed modules, respectively. In addition, compared with other state-of-the-art models, our method can achieve similar or even better classification results, yet offer superior inference performance.
format Article
id doaj-art-c8046fbb9cf8459c95e36a38b5633f1e
institution DOAJ
issn 1939-1404
2151-1535
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-c8046fbb9cf8459c95e36a38b5633f1e2025-08-20T02:55:58ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352024-01-01171921193710.1109/JSTARS.2023.334245310356620Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover ClassificationZongqian Zhan0https://orcid.org/0000-0002-1324-8430Zirou Xiong1Xin Huang2Chun Yang3https://orcid.org/0000-0001-5262-3048Yi Liu4https://orcid.org/0000-0002-6741-6815Xin Wang5https://orcid.org/0000-0002-5132-4465School of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaSchool of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaSchool of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaHangzhou SensingX Technology Company Ltd., Hangzhou, ChinaSchool of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaSchool of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaLand cover classification has the goal to attribute each pixel of high-resolution remoste sensing image with planimetric category labels (such as vegetation, building, and water). In recent years, many serial deep-learning architectures (features are delivered through a single path, such as in <italic>ResNet</italic>, <italic>MobileNet</italic>, and <italic>Segformer</italic>) based on convolutional neural networks and attention mechanisms have been widely explored in land cover classification. However, high-resolution remote sensing images typically have abundant textual details, variable scales in objects, large intraclass variance, and similar interclass correlation, which bring challenges to land cover classification. In this work, we present two pluggable modules to further boost serial learning architecture: first, to cope with ambiguous boundaries caused by lost details and fragmented segmentation stemmed from scale variances, a combination of spatial attention and channel attention is proposed for multiscale feature reconstruction (MSFR); second, to mitigate the classification error caused by intraclass variance and interclass correlation, we explore an interclass attention weighting (ICAW) module, which builds feature vectors for each category, and applies a multihead attention model to capture self-attention dependence among different categories. The experimental results demonstrate that the proposed modules are feasible to the existing serial learning architectures and can improve overall accuracy (OA) by 5.64&#x0025; on the ISPRS Vaihingen two-dimensional dataset (using <italic>ResNet50</italic> as a backbone); in particular, the OA values are 80.68&#x0025; and 86.32&#x0025; before and after using the proposed modules, respectively. In addition, compared with other state-of-the-art models, our method can achieve similar or even better classification results, yet offer superior inference performance.https://ieeexplore.ieee.org/document/10356620/Interclass attentionland cover classificationmultiscale feature reconstruction (MSFR)remote sensing imagesemantic segmentation
spellingShingle Zongqian Zhan
Zirou Xiong
Xin Huang
Chun Yang
Yi Liu
Xin Wang
Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover Classification
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Interclass attention
land cover classification
multiscale feature reconstruction (MSFR)
remote sensing image
semantic segmentation
title Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover Classification
title_full Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover Classification
title_fullStr Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover Classification
title_full_unstemmed Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover Classification
title_short Multiscale Feature Reconstruction and Interclass Attention Weighting for Land Cover Classification
title_sort multiscale feature reconstruction and interclass attention weighting for land cover classification
topic Interclass attention
land cover classification
multiscale feature reconstruction (MSFR)
remote sensing image
semantic segmentation
url https://ieeexplore.ieee.org/document/10356620/
work_keys_str_mv AT zongqianzhan multiscalefeaturereconstructionandinterclassattentionweightingforlandcoverclassification
AT zirouxiong multiscalefeaturereconstructionandinterclassattentionweightingforlandcoverclassification
AT xinhuang multiscalefeaturereconstructionandinterclassattentionweightingforlandcoverclassification
AT chunyang multiscalefeaturereconstructionandinterclassattentionweightingforlandcoverclassification
AT yiliu multiscalefeaturereconstructionandinterclassattentionweightingforlandcoverclassification
AT xinwang multiscalefeaturereconstructionandinterclassattentionweightingforlandcoverclassification