Transformer-based Semantic Segmentation for Large-Scale Building Footprint Extraction from Very-High Resolution Satellite Images

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26220%2F24%3APU151061" target="_blank" >RIV/00216305:26220/24:PU151061 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.sciencedirect.com/science/article/pii/S0273117724002205" target="_blank" >https://www.sciencedirect.com/science/article/pii/S0273117724002205</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.asr.2024.03.002" target="_blank" >10.1016/j.asr.2024.03.002</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Transformer-based Semantic Segmentation for Large-Scale Building Footprint Extraction from Very-High Resolution Satellite Images

Popis výsledku v původním jazyce

Extracting building footprints from extensive very-high spatial resolution (VHSR) remote sensing data is crucial for diverse applications, including surveying, urban studies, population estimation, identification of informal settlements, and disaster management. Although convolutional neural networks (CNNs) are commonly utilized for this purpose, their effectiveness is constrained by limitations in capturing long-range relationships and contextual details due to the localized nature of convolution operations. This study introduces the masked-attention mask transformer (Mask2Former), based on the Swin Transformer, for building footprint extraction from large-scale satellite imagery. To enhance the capture of large-scale semantic information and extract multiscale features, a hierarchical vision transformer with shifted windows (Swin Transformer) serves as the backbone network. An extensive analysis compares the efficiency and generalizability of Mask2Former with four CNN models (PSPNet, DeepLabV3+, UpperNet-ConvNext, and SegNeXt) and two transformer-based models (UpperNet-Swin and SegFormer) featuring different complexities. Results reveal superior performance of transformer-based models over CNN-based counterparts, showcasing exceptional generalization across diverse testing areas with varying building structures, heights, and sizes. Specifically, Mask2Former with the Swin transformer backbone achieves a mean intersection over union between 88% and 93%, along with a mean F-score (mF-score) ranging from 91% to 96.35% across various urban landscapes.

Název v anglickém jazyce

Transformer-based Semantic Segmentation for Large-Scale Building Footprint Extraction from Very-High Resolution Satellite Images

Popis výsledku anglicky

Extracting building footprints from extensive very-high spatial resolution (VHSR) remote sensing data is crucial for diverse applications, including surveying, urban studies, population estimation, identification of informal settlements, and disaster management. Although convolutional neural networks (CNNs) are commonly utilized for this purpose, their effectiveness is constrained by limitations in capturing long-range relationships and contextual details due to the localized nature of convolution operations. This study introduces the masked-attention mask transformer (Mask2Former), based on the Swin Transformer, for building footprint extraction from large-scale satellite imagery. To enhance the capture of large-scale semantic information and extract multiscale features, a hierarchical vision transformer with shifted windows (Swin Transformer) serves as the backbone network. An extensive analysis compares the efficiency and generalizability of Mask2Former with four CNN models (PSPNet, DeepLabV3+, UpperNet-ConvNext, and SegNeXt) and two transformer-based models (UpperNet-Swin and SegFormer) featuring different complexities. Results reveal superior performance of transformer-based models over CNN-based counterparts, showcasing exceptional generalization across diverse testing areas with varying building structures, heights, and sizes. Specifically, Mask2Former with the Swin transformer backbone achieves a mean intersection over union between 88% and 93%, along with a mean F-score (mF-score) ranging from 91% to 96.35% across various urban landscapes.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

20202 - Communication engineering and systems

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

ADVANCES IN SPACE RESEARCH

ISSN

0273-1177

e-ISSN

1879-1948

Svazek periodika

73

Číslo periodika v rámci svazku

10

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

17

Strana od-do

„4937 “-„4954“

Kód UT WoS článku

001226582700001

EID výsledku v databázi Scopus

2-s2.0-85188559920