Transformer-Based Semantic Segmentation for Extraction of Building Footprints from Very-High-Resolution Images
Sensors · 29.05.2023 · 23 (11) · 5166
DOI: 10.3390/s23115166 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 29.05.2023. Verfasst von Jia Song und 2 weiteren Autor:innen. Beteiligte Einrichtung: State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Acad…. Hauptthemen laut MEDLINE: Semantics, Electric Power Supplies. Die Arbeit wurde bisher 18-mal zitiert (Crossref) und nennt 69 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the 14th Five-year Informatization Plan of Chinese Academy of Sciences, National Key Research and Development Program of China, 14th Five-year Informatization Plan of Chinese Academy of Sciences.
Abstract (Originalsprache)
Semantic segmentation with deep learning networks has become an important approach to the extraction of objects from very high-resolution remote sensing images. Vision Transformer networks have shown significant improvements in performance compared to traditional convolutional neural networks (CNNs) in semantic segmentation. Vision Transformer networks have different architectures to CNNs. Image patches, linear embedding, and multi-head self-attention (MHSA) are several of the main hyperparameters. How we should configure them for the extraction of objects in VHR images and how they affect the accuracy of networks are topics that have not been sufficiently investigated. This article explores the role of vision Transformer networks in the extraction of building footprints from very-high-resolution (VHR) images. Transformer-based models with different hyperparameter values were designed and compared, and their impact on accuracy was analyzed. The results show that smaller image patches and higher-dimension embeddings result in better accuracy. In addition, the Transformer-based network is shown to be scalable and can be trained with general-scale graphics processing units (GPUs) with comparable model sizes and training times to convolutional neural networks while achieving higher accuracy. The study provides valuable insights into the potential of vision Transformer networks in object extraction using VHR images.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China.
- Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application, Nanjing 210023, China.
- Department of Geography, University of Wisconsin, Madison, WI 53706, USA.
ORCID-Profile: Jia Song
Förderung
- the 14th Five-year Informatization Plan of Chinese Academy of Sciences
- National Key Research and Development Program of China
- 14th Five-year Informatization Plan of Chinese Academy of Sciences
Zitieren
APA
Song, J., Zhu, A. X., & Zhu, Y. (2023). Transformer-Based Semantic Segmentation for Extraction of Building Footprints from Very-High-Resolution Images. Sensors, 23(11), 5166. https://doi.org/10.3390/s23115166
Vancouver
Song J, Zhu AX, Zhu Y. Transformer-Based Semantic Segmentation for Extraction of Building Footprints from Very-High-Resolution Images. Sensors. 2023;23(11):5166. doi:10.3390/s23115166
BibTeX
@article{Song2023,
author = {Song, Jia and Zhu, A-Xing and Zhu, Yunqiang},
title = {Transformer-Based Semantic Segmentation for Extraction of Building Footprints from Very-High-Resolution Images},
journal = {Sensors},
year = {2023},
volume = {23},
number = {11},
pages = {5166},
doi = {10.3390/s23115166}
}
EUVIMED