FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and Analysis

Semantic segmentation of vegetation in aerial remote sensing images is a critical aspect of vegetation mapping. Accurate vegetation segmentation effectively informs real-world production and construction activities. However, the presence of species heterogeneity, seasonal variations, and feature dis...

Full description

Saved in:
Bibliographic Details
Main Authors: Bingnan He, Dongyang Wu, Li Wang, Sheng Xu
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/16/22/4194
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850266750487101440
author Bingnan He
Dongyang Wu
Li Wang
Sheng Xu
author_facet Bingnan He
Dongyang Wu
Li Wang
Sheng Xu
author_sort Bingnan He
collection DOAJ
description Semantic segmentation of vegetation in aerial remote sensing images is a critical aspect of vegetation mapping. Accurate vegetation segmentation effectively informs real-world production and construction activities. However, the presence of species heterogeneity, seasonal variations, and feature disparities within remote sensing images poses significant challenges for vision tasks. Traditional machine learning-based methods often struggle to capture deep-level features for the segmentation. This work proposes a novel deep learning network named FA-HRNet that leverages the fusion of attention mechanism and a multi-branch network structure for vegetation detection and segmentation. Quantitative analysis from multiple datasets reveals that our method outperforms existing approaches, with improvements in MIoU and PA by 2.17% and 4.85%, respectively, compared with the baseline network. Our approach exhibits significant advantages over the other methods regarding cross-region and cross-scale capabilities, providing a reliable vegetation coverage ratio for ecological analysis.
format Article
id doaj-art-226251b1d7c8419d8154e70fd7779e54
institution OA Journals
issn 2072-4292
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-226251b1d7c8419d8154e70fd7779e542025-08-20T01:54:04ZengMDPI AGRemote Sensing2072-42922024-11-011622419410.3390/rs16224194FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and AnalysisBingnan He0Dongyang Wu1Li Wang2Sheng Xu3College of Information Science and Technology and Artificial Intelligence, Nanjing Forestry University, Nanjing 210037, ChinaCollege of Information Science and Technology and Artificial Intelligence, Nanjing Forestry University, Nanjing 210037, ChinaCollege of Information Science and Technology and Artificial Intelligence, Nanjing Forestry University, Nanjing 210037, ChinaCollege of Information Science and Technology and Artificial Intelligence, Nanjing Forestry University, Nanjing 210037, ChinaSemantic segmentation of vegetation in aerial remote sensing images is a critical aspect of vegetation mapping. Accurate vegetation segmentation effectively informs real-world production and construction activities. However, the presence of species heterogeneity, seasonal variations, and feature disparities within remote sensing images poses significant challenges for vision tasks. Traditional machine learning-based methods often struggle to capture deep-level features for the segmentation. This work proposes a novel deep learning network named FA-HRNet that leverages the fusion of attention mechanism and a multi-branch network structure for vegetation detection and segmentation. Quantitative analysis from multiple datasets reveals that our method outperforms existing approaches, with improvements in MIoU and PA by 2.17% and 4.85%, respectively, compared with the baseline network. Our approach exhibits significant advantages over the other methods regarding cross-region and cross-scale capabilities, providing a reliable vegetation coverage ratio for ecological analysis.https://www.mdpi.com/2072-4292/16/22/4194computer visionaerial remote sensing imagesvision modelssemantic segmentationvegetation coveragepanoptic segmentation
spellingShingle Bingnan He
Dongyang Wu
Li Wang
Sheng Xu
FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and Analysis
Remote Sensing
computer vision
aerial remote sensing images
vision models
semantic segmentation
vegetation coverage
panoptic segmentation
title FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and Analysis
title_full FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and Analysis
title_fullStr FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and Analysis
title_full_unstemmed FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and Analysis
title_short FA-HRNet: A New Fusion Attention Approach for Vegetation Semantic Segmentation and Analysis
title_sort fa hrnet a new fusion attention approach for vegetation semantic segmentation and analysis
topic computer vision
aerial remote sensing images
vision models
semantic segmentation
vegetation coverage
panoptic segmentation
url https://www.mdpi.com/2072-4292/16/22/4194
work_keys_str_mv AT bingnanhe fahrnetanewfusionattentionapproachforvegetationsemanticsegmentationandanalysis
AT dongyangwu fahrnetanewfusionattentionapproachforvegetationsemanticsegmentationandanalysis
AT liwang fahrnetanewfusionattentionapproachforvegetationsemanticsegmentationandanalysis
AT shengxu fahrnetanewfusionattentionapproachforvegetationsemanticsegmentationandanalysis