Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery

Abstract Accurate landslide segmentation using remote sensing imagery is a critical component of geohazards response systems, particularly in time-sensitive tasks such as post-earthquake landslide damage assessment and emergency resource allocation. However, current methodologies struggle with two p...

Full description

Saved in:
Bibliographic Details
Main Authors: Wenjie Hu, Guangtong Sun, Xiangqiang Zeng, Bo Tong, Zihao Wang, Xinyue Wu, Ping Song
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-08695-8
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849769439292030976
author Wenjie Hu
Guangtong Sun
Xiangqiang Zeng
Bo Tong
Zihao Wang
Xinyue Wu
Ping Song
author_facet Wenjie Hu
Guangtong Sun
Xiangqiang Zeng
Bo Tong
Zihao Wang
Xinyue Wu
Ping Song
author_sort Wenjie Hu
collection DOAJ
description Abstract Accurate landslide segmentation using remote sensing imagery is a critical component of geohazards response systems, particularly in time-sensitive tasks such as post-earthquake landslide damage assessment and emergency resource allocation. However, current methodologies struggle with two persistent challenges in sub-meter true-color imagery: fine-grained inter-class confusion between landslides and spectrally analogous terrain features, and within-landslide heterogeneity where localized damage signatures coexist with macro-scale deformation patterns within individual landslide bodies. To overcome these, we propose the Cross-Attention Landslide Detector (CALandDet), which improves the model’s ability to distinguish between landslide and background features by sharply capturing global landslide feature information and integrating global landslide feature information with local information via a cross-attention feature enhancement mechanism. Ablation experiments show that CALandDet outperforms baselines, as evidenced by a 4.89% enhanced F1 score and an 8.73% greater Intersection over Union (IoU). In comparative experiments, it outperforms the other models by 8.05–10.78% in IoU and 1.05–8.9% in F1 score, achieving an IoU of 82.65% and an F1 score of 81.64%. Furthermore, the Gradient-weighted Class Activation Mapping (Grad-CAM) visualizations confirm that the decision regions generated by the CALandDet model exhibit a higher spatial consistency with the actual landslide areas, effectively capturing indicative features including surface textures, sliding debris, accumulation bodies, and vegetation destruction. The proposed method may serve as a reference for future advancements in landslide segmentation and other remote sensing segmentation tasks.
format Article
id doaj-art-af02edacbc714f02a86a7708bb109a6e
institution DOAJ
issn 2045-2322
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-af02edacbc714f02a86a7708bb109a6e2025-08-20T03:03:24ZengNature PortfolioScientific Reports2045-23222025-07-0115111710.1038/s41598-025-08695-8Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imageryWenjie Hu0Guangtong Sun1Xiangqiang Zeng2Bo Tong3Zihao Wang4Xinyue Wu5Ping Song6Institute of Disaster PreventionInstitute of Disaster PreventionState Key Laboratory of Remote Sensing Science, Faculty of Geographical Science, Beijing Normal UniversityNorth China Institute of Science and TechnologyCollege of Physics and Optoelectronic Engineering, Ocean University of ChinaInstitute of Disaster PreventionInstitute of Disaster PreventionAbstract Accurate landslide segmentation using remote sensing imagery is a critical component of geohazards response systems, particularly in time-sensitive tasks such as post-earthquake landslide damage assessment and emergency resource allocation. However, current methodologies struggle with two persistent challenges in sub-meter true-color imagery: fine-grained inter-class confusion between landslides and spectrally analogous terrain features, and within-landslide heterogeneity where localized damage signatures coexist with macro-scale deformation patterns within individual landslide bodies. To overcome these, we propose the Cross-Attention Landslide Detector (CALandDet), which improves the model’s ability to distinguish between landslide and background features by sharply capturing global landslide feature information and integrating global landslide feature information with local information via a cross-attention feature enhancement mechanism. Ablation experiments show that CALandDet outperforms baselines, as evidenced by a 4.89% enhanced F1 score and an 8.73% greater Intersection over Union (IoU). In comparative experiments, it outperforms the other models by 8.05–10.78% in IoU and 1.05–8.9% in F1 score, achieving an IoU of 82.65% and an F1 score of 81.64%. Furthermore, the Gradient-weighted Class Activation Mapping (Grad-CAM) visualizations confirm that the decision regions generated by the CALandDet model exhibit a higher spatial consistency with the actual landslide areas, effectively capturing indicative features including surface textures, sliding debris, accumulation bodies, and vegetation destruction. The proposed method may serve as a reference for future advancements in landslide segmentation and other remote sensing segmentation tasks.https://doi.org/10.1038/s41598-025-08695-8Cross-attentionLandslide segmentationDeep learningLandslideGeohazards
spellingShingle Wenjie Hu
Guangtong Sun
Xiangqiang Zeng
Bo Tong
Zihao Wang
Xinyue Wu
Ping Song
Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery
Scientific Reports
Cross-attention
Landslide segmentation
Deep learning
Landslide
Geohazards
title Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery
title_full Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery
title_fullStr Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery
title_full_unstemmed Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery
title_short Hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery
title_sort hierarchical cross attention achieves pixel precise landslide segmentation in submeter optical imagery
topic Cross-attention
Landslide segmentation
Deep learning
Landslide
Geohazards
url https://doi.org/10.1038/s41598-025-08695-8
work_keys_str_mv AT wenjiehu hierarchicalcrossattentionachievespixelpreciselandslidesegmentationinsubmeteropticalimagery
AT guangtongsun hierarchicalcrossattentionachievespixelpreciselandslidesegmentationinsubmeteropticalimagery
AT xiangqiangzeng hierarchicalcrossattentionachievespixelpreciselandslidesegmentationinsubmeteropticalimagery
AT botong hierarchicalcrossattentionachievespixelpreciselandslidesegmentationinsubmeteropticalimagery
AT zihaowang hierarchicalcrossattentionachievespixelpreciselandslidesegmentationinsubmeteropticalimagery
AT xinyuewu hierarchicalcrossattentionachievespixelpreciselandslidesegmentationinsubmeteropticalimagery
AT pingsong hierarchicalcrossattentionachievespixelpreciselandslidesegmentationinsubmeteropticalimagery