Next Article in Journal
Effects of Assimilating Clear-Sky FY-3D MWHS2 Radiance on the Numerical Simulation of Tropical Storm Ampil
Next Article in Special Issue
Learning Adjustable Reduced Downsampling Network for Small Object Detection in Urban Environments
Previous Article in Journal
Preface: Earth Observations for Environmental Sustainability for the Next Decade
Article

Making Low-Resolution Satellite Images Reborn: A Deep Learning Approach for Super-Resolution Building Extraction

1
Department of Earth System Science, Tsinghua University, Beijing 100084, China
2
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China
3
CUHK-SenseTime Joint Lab, The Chinese University of Hong Kong, Hong Kong, China
*
Author to whom correspondence should be addressed.
Academic Editor: Lefei Zhang
Remote Sens. 2021, 13(15), 2872; https://0-doi-org.brum.beds.ac.uk/10.3390/rs13152872
Received: 5 July 2021 / Revised: 18 July 2021 / Accepted: 20 July 2021 / Published: 22 July 2021
(This article belongs to the Special Issue Deep Learning in Remote Sensing Application)
Existing methods for building extraction from remotely sensed images strongly rely on aerial or satellite-based images with very high resolution, which are usually limited by spatiotemporally accessibility and cost. In contrast, relatively low-resolution images have better spatial and temporal availability but cannot directly contribute to fine- and/or high-resolution building extraction. In this paper, based on image super-resolution and segmentation techniques, we propose a two-stage framework (SRBuildingSeg) for achieving super-resolution (SR) building extraction using relatively low-resolution remotely sensed images. SRBuildingSeg can fully utilize inherent information from the given low-resolution images to achieve high-resolution building extraction. In contrast to the existing building extraction methods, we first utilize an internal pairs generation module (IPG) to obtain SR training datasets from the given low-resolution images and an edge-aware super-resolution module (EASR) to improve the perceptional features, following the dual-encoder building segmentation module (DES). Both qualitative and quantitative experimental results demonstrate that our proposed approach is capable of achieving high-resolution (e.g., 0.5 m) building extraction results at 2×, 4× and 8× SR. Our approach outperforms eight other methods with respect to the extraction result of mean Intersection over Union (mIoU) values by a ratio of 9.38%, 8.20%, and 7.89% with SR ratio factors of 2, 4, and 8, respectively. The results indicate that the edges and borders reconstructed in super-resolved images serve a pivotal role in subsequent building extraction and reveal the potential of the proposed approach to achieve super-resolution building extraction. View Full-Text
Keywords: remote sensing imagery; building extraction; super-resolution; deep learning remote sensing imagery; building extraction; super-resolution; deep learning
Show Figures

Graphical abstract

MDPI and ACS Style

Zhang, L.; Dong, R.; Yuan, S.; Li, W.; Zheng, J.; Fu, H. Making Low-Resolution Satellite Images Reborn: A Deep Learning Approach for Super-Resolution Building Extraction. Remote Sens. 2021, 13, 2872. https://0-doi-org.brum.beds.ac.uk/10.3390/rs13152872

AMA Style

Zhang L, Dong R, Yuan S, Li W, Zheng J, Fu H. Making Low-Resolution Satellite Images Reborn: A Deep Learning Approach for Super-Resolution Building Extraction. Remote Sensing. 2021; 13(15):2872. https://0-doi-org.brum.beds.ac.uk/10.3390/rs13152872

Chicago/Turabian Style

Zhang, Lixian, Runmin Dong, Shuai Yuan, Weijia Li, Juepeng Zheng, and Haohuan Fu. 2021. "Making Low-Resolution Satellite Images Reborn: A Deep Learning Approach for Super-Resolution Building Extraction" Remote Sensing 13, no. 15: 2872. https://0-doi-org.brum.beds.ac.uk/10.3390/rs13152872

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop