DESAT: A Distance-Enhanced Strip Attention Transformer for Remote Sensing Image Super-Resolution DOI Creative Commons
Yujie Mao, Guojin He, Guizhou Wang

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(22), P. 4251 - 4251

Published: Nov. 14, 2024

Transformer-based methods have demonstrated impressive performance in image super-resolution tasks. However, when applied to large-scale Earth observation images, the existing transformers encounter two significant challenges: (1) insufficient consideration of spatial correlation between adjacent ground objects; and (2) bottlenecks due underutilization upsample module. To address these issues, we propose a novel distance-enhanced strip attention transformer (DESAT). The DESAT integrates distance priors, easily obtainable from remote sensing into window self-attention mechanism capture correlations more effectively. further enhance transfer deep features high-resolution outputs, designed an attention-enhanced block, which combines pixel shuffle layer with attention-based branch implemented through overlapping mechanism. Additionally, better simulate real-world scenarios, constructed new cross-sensor dataset using Gaofen-6 satellite imagery. Extensive experiments on both simulated datasets demonstrate that outperforms state-of-the-art models by up 1.17 dB along superior qualitative results. Furthermore, achieves competitive tasks, effectively balancing detail reconstruction spectral transform, making it highly suitable for practical applications.

Language: Английский

A Segment Anything Model based weakly supervised learning method for crop mapping using Sentinel-2 time series images DOI Creative Commons
Jialin Sun, Shuai Yan, Xiaochuang Yao

et al.

International Journal of Applied Earth Observation and Geoinformation, Journal Year: 2024, Volume and Issue: 133, P. 104085 - 104085

Published: Aug. 10, 2024

Language: Английский

Citations

7
Satellite images reveal rapid development of global water-based photovoltaic over the past 20 years DOI Creative Commons

He Ren,

Zhen Yang, Fashuai Li

et al.

International Journal of Applied Earth Observation and Geoinformation, Journal Year: 2025, Volume and Issue: 136, P. 104354 - 104354

Published: Jan. 10, 2025

Language: Английский

Citations

0
DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection DOI Creative Commons
Shaofu Lin, Yang Yang, Xiliang Liu

et al.

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(2), P. 332 - 332

Published: Jan. 18, 2025

Precise statistics on the spatial distribution of photovoltaics (PV) are essential for advancing PV industry, and integrating remote sensing with artificial intelligence technologies offers a robust solution accurate identification. Currently, numerous studies focus detection single-type installations through aerial or satellite imagery. However, due to variability in scale shape complex environments, results often fail capture detailed information struggle multi-scale systems. To tackle these challenges, method known as Dynamic Spatial-Frequency Attention SwinNet (DSFA-SwinNet) areas is proposed. First, this study proposes (DSFA) mechanism, Pyramid Refinement (PAR) bottleneck structure, optimizes feature propagation achieve dynamic decoupling frequency domains representation learning. Secondly, hybrid loss function has been developed weights optimized employing Bayesian Optimization algorithm provide strategic parameter tuning similar research. Lastly, fixed window size Swin-Transformer dynamically adjusted enhance computational efficiency maintain accuracy. The two datasets demonstrate that DSFA-SwinNet significantly enhances accuracy scalability areas.

Language: Английский

Citations

0
A large-scale ultra-high-resolution segmentation dataset augmentation framework for photovoltaic panels in photovoltaic power plants based on priori knowledge DOI
Ruiqing Yang, Guojin He, Ranyu Yin

et al.

Applied Energy, Journal Year: 2025, Volume and Issue: 390, P. 125879 - 125879

Published: April 10, 2025

Language: Английский

Citations

0
Integrating unsupervised domain adaptation and SAM technologies for image semantic segmentation: a case study on building extraction from high-resolution remote sensing images DOI Creative Commons

Mengyuan Yang,

Rui Yang, Min Wang

et al.

International Journal of Digital Earth, Journal Year: 2025, Volume and Issue: 18(1)

Published: April 15, 2025

Language: Английский

Citations

0
Uncovering the Location of Photovoltaic Power Plants Using Heterogeneous Remote Sensing Imagery DOI Creative Commons
Siyuan Wang, Bowen Cai, Dongyang Hou

et al.

Energy and AI, Journal Year: 2025, Volume and Issue: unknown, P. 100527 - 100527

Published: May 1, 2025

Language: Английский

Citations

0
Toward global rooftop PV detection with Deep Active Learning DOI Creative Commons
Matthias Zech,

Hendrik-Pieter Tetens,

Joseph Ranalli

et al.

Advances in Applied Energy, Journal Year: 2024, Volume and Issue: unknown, P. 100191 - 100191

Published: Sept. 1, 2024

Language: Английский

Citations

2
DESAT: A Distance-Enhanced Strip Attention Transformer for Remote Sensing Image Super-Resolution DOI Creative Commons
Yujie Mao, Guojin He, Guizhou Wang

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(22), P. 4251 - 4251

Published: Nov. 14, 2024

Transformer-based methods have demonstrated impressive performance in image super-resolution tasks. However, when applied to large-scale Earth observation images, the existing transformers encounter two significant challenges: (1) insufficient consideration of spatial correlation between adjacent ground objects; and (2) bottlenecks due underutilization upsample module. To address these issues, we propose a novel distance-enhanced strip attention transformer (DESAT). The DESAT integrates distance priors, easily obtainable from remote sensing into window self-attention mechanism capture correlations more effectively. further enhance transfer deep features high-resolution outputs, designed an attention-enhanced block, which combines pixel shuffle layer with attention-based branch implemented through overlapping mechanism. Additionally, better simulate real-world scenarios, constructed new cross-sensor dataset using Gaofen-6 satellite imagery. Extensive experiments on both simulated datasets demonstrate that outperforms state-of-the-art models by up 1.17 dB along superior qualitative results. Furthermore, achieves competitive tasks, effectively balancing detail reconstruction spectral transform, making it highly suitable for practical applications.

Language: Английский

Citations

0