Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation Mamba

The semantic segmentation of satellite and UAV remote sensing imagery is pivotal for address exploration, change detection, quantitative analysis and urban planning. Recent advancements have seen an influx of segmentation networks utilizing convolutional neural networks and transformers. However, th...

Full description

Saved in:
Bibliographic Details
Main Authors: Qi Zhang, Guohua Geng, Pengbo Zhou, Qinglin Liu, Yong Wang, Kang Li
Format: Article
Language:English
Published: MDPI AG 2024-09-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/16/19/3622
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850284522430529536
author Qi Zhang
Guohua Geng
Pengbo Zhou
Qinglin Liu
Yong Wang
Kang Li
author_facet Qi Zhang
Guohua Geng
Pengbo Zhou
Qinglin Liu
Yong Wang
Kang Li
author_sort Qi Zhang
collection DOAJ
description The semantic segmentation of satellite and UAV remote sensing imagery is pivotal for address exploration, change detection, quantitative analysis and urban planning. Recent advancements have seen an influx of segmentation networks utilizing convolutional neural networks and transformers. However, the intricate geographical features and varied land cover boundary interferences in remote sensing imagery still challenge conventional segmentation networks’ spatial representation and long-range dependency capabilities. This paper introduces a novel U-Net-like network for UAV image segmentation. We developed a link aggregation Mamba at the critical skip connection stage of UNetFormer. This approach maps and aggregates multi-scale features from different stages into a unified linear dimension through four Mamba branches containing state-space models (SSMs), ultimately decoupling and fusing these features to restore the contextual relationships in the mask. Moreover, the Mix-Mamba module is incorporated, leveraging a parallel self-attention mechanism with SSMs to merge the advantages of a global receptive field and reduce modeling complexity. This module facilitates nonlinear modeling across different channels and spaces through multipath activation, catering to international and local long-range dependencies. Evaluations on public remote sensing datasets like LovaDA, UAVid and Vaihingen underscore the state-of-the-art performance of our approach.
format Article
id doaj-art-10e508a2a220490d9fbb5140af3cb52e
institution OA Journals
issn 2072-4292
language English
publishDate 2024-09-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-10e508a2a220490d9fbb5140af3cb52e2025-08-20T01:47:33ZengMDPI AGRemote Sensing2072-42922024-09-011619362210.3390/rs16193622Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation MambaQi Zhang0Guohua Geng1Pengbo Zhou2Qinglin Liu3Yong Wang4Kang Li5School of Information Science and Technology, Northwest University, Xi’an 710127, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710127, ChinaSchool of Arts and Communication, Beijing Normal University, Beijing 100875, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710127, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710127, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710127, ChinaThe semantic segmentation of satellite and UAV remote sensing imagery is pivotal for address exploration, change detection, quantitative analysis and urban planning. Recent advancements have seen an influx of segmentation networks utilizing convolutional neural networks and transformers. However, the intricate geographical features and varied land cover boundary interferences in remote sensing imagery still challenge conventional segmentation networks’ spatial representation and long-range dependency capabilities. This paper introduces a novel U-Net-like network for UAV image segmentation. We developed a link aggregation Mamba at the critical skip connection stage of UNetFormer. This approach maps and aggregates multi-scale features from different stages into a unified linear dimension through four Mamba branches containing state-space models (SSMs), ultimately decoupling and fusing these features to restore the contextual relationships in the mask. Moreover, the Mix-Mamba module is incorporated, leveraging a parallel self-attention mechanism with SSMs to merge the advantages of a global receptive field and reduce modeling complexity. This module facilitates nonlinear modeling across different channels and spaces through multipath activation, catering to international and local long-range dependencies. Evaluations on public remote sensing datasets like LovaDA, UAVid and Vaihingen underscore the state-of-the-art performance of our approach.https://www.mdpi.com/2072-4292/16/19/3622semantic segmentationremote sensingMambastate-space modellink aggregation
spellingShingle Qi Zhang
Guohua Geng
Pengbo Zhou
Qinglin Liu
Yong Wang
Kang Li
Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation Mamba
Remote Sensing
semantic segmentation
remote sensing
Mamba
state-space model
link aggregation
title Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation Mamba
title_full Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation Mamba
title_fullStr Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation Mamba
title_full_unstemmed Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation Mamba
title_short Link Aggregation for Skip Connection–Mamba: Remote Sensing Image Segmentation Network Based on Link Aggregation Mamba
title_sort link aggregation for skip connection mamba remote sensing image segmentation network based on link aggregation mamba
topic semantic segmentation
remote sensing
Mamba
state-space model
link aggregation
url https://www.mdpi.com/2072-4292/16/19/3622
work_keys_str_mv AT qizhang linkaggregationforskipconnectionmambaremotesensingimagesegmentationnetworkbasedonlinkaggregationmamba
AT guohuageng linkaggregationforskipconnectionmambaremotesensingimagesegmentationnetworkbasedonlinkaggregationmamba
AT pengbozhou linkaggregationforskipconnectionmambaremotesensingimagesegmentationnetworkbasedonlinkaggregationmamba
AT qinglinliu linkaggregationforskipconnectionmambaremotesensingimagesegmentationnetworkbasedonlinkaggregationmamba
AT yongwang linkaggregationforskipconnectionmambaremotesensingimagesegmentationnetworkbasedonlinkaggregationmamba
AT kangli linkaggregationforskipconnectionmambaremotesensingimagesegmentationnetworkbasedonlinkaggregationmamba