AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing Images

The automatic extraction of buildings from remote sensing images is crucial for various applications such as urban planning and management, emergency response, and map making and updating. In recent years, deep learning (DL) methods have made significant progress in this field. However, due to the c...

Full description

Saved in:
Bibliographic Details
Main Authors: Bowen Zhu, Ding Yu, Xiongwu Xiao, Jian Shen, Zhigao Cui, Yanzhao Su, Aihua Li, Deren Li
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/17/9/1481
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850030748001632256
author Bowen Zhu
Ding Yu
Xiongwu Xiao
Jian Shen
Zhigao Cui
Yanzhao Su
Aihua Li
Deren Li
author_facet Bowen Zhu
Ding Yu
Xiongwu Xiao
Jian Shen
Zhigao Cui
Yanzhao Su
Aihua Li
Deren Li
author_sort Bowen Zhu
collection DOAJ
description The automatic extraction of buildings from remote sensing images is crucial for various applications such as urban planning and management, emergency response, and map making and updating. In recent years, deep learning (DL) methods have made significant progress in this field. However, due to the complex and diverse structures of buildings and their interconnections, the accuracy of extracted buildings remains insufficient for high-precision applications such as maps and navigation. To address the issue of enhancing building boundary extraction, we propose a modified instance segmentation model, AP-PointRend (Adaptive Parameter-PointRend), to improve the performance of building instance extraction. Specifically, the model can adaptively select the number of iterations and points based on the size of buildings to improve the segmentation accuracy of large buildings. By introducing regularization constraints, discrete small patches are removed, preserving boundaries better during the segmentation process. We also designed an image merging method to eliminate seams, ensure the recall rate, and improve the extraction accuracy. The Vaihingen and WHU benchmark datasets were used to evaluate the performance of the AP-PointRend method. The experimental results showed that the proposed AP-PointRend approach generated better building extraction results compared with other state-of-the-art methods.
format Article
id doaj-art-a9015b2da6df4c0c994c1e4f9109fa79
institution DOAJ
issn 2072-4292
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-a9015b2da6df4c0c994c1e4f9109fa792025-08-20T02:59:08ZengMDPI AGRemote Sensing2072-42922025-04-01179148110.3390/rs17091481AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing ImagesBowen Zhu0Ding Yu1Xiongwu Xiao2Jian Shen3Zhigao Cui4Yanzhao Su5Aihua Li6Deren Li7School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430072, ChinaKey Laboratory of the Ministry of Education on Application of Artificial Intelligence in Equipment, Xi’an Research Institute of High Technology, Xi’an 710025, ChinaState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing (LIESMARS), Wuhan University, Wuhan 430072, ChinaSchool of Computer Science, Wuhan University, Wuhan 430072, ChinaKey Laboratory of the Ministry of Education on Application of Artificial Intelligence in Equipment, Xi’an Research Institute of High Technology, Xi’an 710025, ChinaKey Laboratory of the Ministry of Education on Application of Artificial Intelligence in Equipment, Xi’an Research Institute of High Technology, Xi’an 710025, ChinaKey Laboratory of the Ministry of Education on Application of Artificial Intelligence in Equipment, Xi’an Research Institute of High Technology, Xi’an 710025, ChinaState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing (LIESMARS), Wuhan University, Wuhan 430072, ChinaThe automatic extraction of buildings from remote sensing images is crucial for various applications such as urban planning and management, emergency response, and map making and updating. In recent years, deep learning (DL) methods have made significant progress in this field. However, due to the complex and diverse structures of buildings and their interconnections, the accuracy of extracted buildings remains insufficient for high-precision applications such as maps and navigation. To address the issue of enhancing building boundary extraction, we propose a modified instance segmentation model, AP-PointRend (Adaptive Parameter-PointRend), to improve the performance of building instance extraction. Specifically, the model can adaptively select the number of iterations and points based on the size of buildings to improve the segmentation accuracy of large buildings. By introducing regularization constraints, discrete small patches are removed, preserving boundaries better during the segmentation process. We also designed an image merging method to eliminate seams, ensure the recall rate, and improve the extraction accuracy. The Vaihingen and WHU benchmark datasets were used to evaluate the performance of the AP-PointRend method. The experimental results showed that the proposed AP-PointRend approach generated better building extraction results compared with other state-of-the-art methods.https://www.mdpi.com/2072-4292/17/9/1481building extractioninstance segmentationdeep learningremote sensing
spellingShingle Bowen Zhu
Ding Yu
Xiongwu Xiao
Jian Shen
Zhigao Cui
Yanzhao Su
Aihua Li
Deren Li
AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing Images
Remote Sensing
building extraction
instance segmentation
deep learning
remote sensing
title AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing Images
title_full AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing Images
title_fullStr AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing Images
title_full_unstemmed AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing Images
title_short AP-PointRend: An Improved Network for Building Extraction via High-Resolution Remote Sensing Images
title_sort ap pointrend an improved network for building extraction via high resolution remote sensing images
topic building extraction
instance segmentation
deep learning
remote sensing
url https://www.mdpi.com/2072-4292/17/9/1481
work_keys_str_mv AT bowenzhu appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages
AT dingyu appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages
AT xiongwuxiao appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages
AT jianshen appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages
AT zhigaocui appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages
AT yanzhaosu appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages
AT aihuali appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages
AT derenli appointrendanimprovednetworkforbuildingextractionviahighresolutionremotesensingimages