GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR data

Accurate semantic segmentation of airborne LiDAR point clouds is essential for the intelligent inspection and maintenance of high-voltage transmission infrastructure. While existing methods predominantly focus on major structural components such as towers and conductors, they often fail to address t...

Full description

Saved in:
Bibliographic Details
Main Authors: Su Zhang, Haibo Liu, Jingguo Rong, Yaping Zhang
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-08-01
Series:Frontiers in Earth Science
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/feart.2025.1649203/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849231591906213888
author Su Zhang
Haibo Liu
Jingguo Rong
Yaping Zhang
author_facet Su Zhang
Haibo Liu
Jingguo Rong
Yaping Zhang
author_sort Su Zhang
collection DOAJ
description Accurate semantic segmentation of airborne LiDAR point clouds is essential for the intelligent inspection and maintenance of high-voltage transmission infrastructure. While existing methods predominantly focus on major structural components such as towers and conductors, they often fail to address the fine-grained segmentation of smaller yet critical elements, including ground wires, crossing lines, and insulators. To tackle this limitation, we propose a novel network architecture—Graph-Kernel Convolution Attention Encoder (GKCAE)—designed for multi-class, fine-grained semantic segmentation of transmission corridor point clouds. GKCAE first captures local geometric features using Kernel Point Convolution, and then models inter-class spatial relationships through Graph Edge-Conditioned Convolution to incorporate global contextual information. Additionally, a Channel-Spatial Attention Module is introduced to enhance point-level feature representations, particularly for small or geometrically similar classes. Experiments conducted on three realworld transmission corridor datasets demonstrate that our method achieves a mean Intersection over Union (mIoU) of 81.93% and an Overall Accuracy (OA) of 94.1%, outperforming existing state-of-the-art approaches.
format Article
id doaj-art-bf59bb1bcf3b4f7bb17b0eebd8bdd071
institution Kabale University
issn 2296-6463
language English
publishDate 2025-08-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Earth Science
spelling doaj-art-bf59bb1bcf3b4f7bb17b0eebd8bdd0712025-08-21T05:27:28ZengFrontiers Media S.A.Frontiers in Earth Science2296-64632025-08-011310.3389/feart.2025.16492031649203GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR dataSu ZhangHaibo LiuJingguo RongYaping ZhangAccurate semantic segmentation of airborne LiDAR point clouds is essential for the intelligent inspection and maintenance of high-voltage transmission infrastructure. While existing methods predominantly focus on major structural components such as towers and conductors, they often fail to address the fine-grained segmentation of smaller yet critical elements, including ground wires, crossing lines, and insulators. To tackle this limitation, we propose a novel network architecture—Graph-Kernel Convolution Attention Encoder (GKCAE)—designed for multi-class, fine-grained semantic segmentation of transmission corridor point clouds. GKCAE first captures local geometric features using Kernel Point Convolution, and then models inter-class spatial relationships through Graph Edge-Conditioned Convolution to incorporate global contextual information. Additionally, a Channel-Spatial Attention Module is introduced to enhance point-level feature representations, particularly for small or geometrically similar classes. Experiments conducted on three realworld transmission corridor datasets demonstrate that our method achieves a mean Intersection over Union (mIoU) of 81.93% and an Overall Accuracy (OA) of 94.1%, outperforming existing state-of-the-art approaches.https://www.frontiersin.org/articles/10.3389/feart.2025.1649203/fullALS point cloudssemantic segmentationgraph edge convolutionhigh-voltage transmission corridorsdeep learning
spellingShingle Su Zhang
Haibo Liu
Jingguo Rong
Yaping Zhang
GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR data
Frontiers in Earth Science
ALS point clouds
semantic segmentation
graph edge convolution
high-voltage transmission corridors
deep learning
title GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR data
title_full GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR data
title_fullStr GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR data
title_full_unstemmed GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR data
title_short GKCAE: A graph-attention-based encoder for fine-grained semantic segmentation of high-voltage transmission corridors scenario LiDAR data
title_sort gkcae a graph attention based encoder for fine grained semantic segmentation of high voltage transmission corridors scenario lidar data
topic ALS point clouds
semantic segmentation
graph edge convolution
high-voltage transmission corridors
deep learning
url https://www.frontiersin.org/articles/10.3389/feart.2025.1649203/full
work_keys_str_mv AT suzhang gkcaeagraphattentionbasedencoderforfinegrainedsemanticsegmentationofhighvoltagetransmissioncorridorsscenariolidardata
AT haiboliu gkcaeagraphattentionbasedencoderforfinegrainedsemanticsegmentationofhighvoltagetransmissioncorridorsscenariolidardata
AT jingguorong gkcaeagraphattentionbasedencoderforfinegrainedsemanticsegmentationofhighvoltagetransmissioncorridorsscenariolidardata
AT yapingzhang gkcaeagraphattentionbasedencoderforfinegrainedsemanticsegmentationofhighvoltagetransmissioncorridorsscenariolidardata