相关参考文献
注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article
Geochemistry & Geophysics
Gianluca Giuffrida et al.
Summary: Artificial intelligence (AI) is revolutionizing algorithm development and allowing for the direct extraction of relevant information from data. Recent advances in microelectronics have enabled the implementation of AI algorithms at the edge, reducing data bandwidth and latency. The European Space Agency (ESA) has been promoting the development of disruptive technologies for earth observation missions, with Phi-sat-1 being a pioneering experiment showcasing the potential of on-board AI for cloud detection. This mission successfully ran an AI deep convolutional neural network directly on a dedicated accelerator on-board a satellite, marking a new era of discovery and commercial applications.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Environmental Sciences
Iyke Maduako et al.
Summary: Computer vision for large scale building detection is challenging, especially when detecting specific buildings globally. However, using deep learning techniques and high-resolution satellite imagery, a Deep Convolutional Neural Network (CNN) can be trained to successfully map school locations. Regional models perform better due to exposure to diverse school structures and features, while the global model struggles with generalizing across different regions and countries.
Article
Multidisciplinary Sciences
L. Kruitwagen et al.
Summary: The global inventory of utility-scale solar photovoltaic generating units revealed that the majority of facilities are located on cropland, with continuously increasing generating capacity. Additional geospatial data is necessary to manage generation intermittency, mitigate climate change risks. The inventory could potentially assist in aligning PV delivery with the Sustainable Development Goals.
Article
Environmental Sciences
Dan Lopez-Puigdollers et al.
Summary: This study systematically compares deep learning models trained on different Landsat-8 and Sentinel-2 datasets, showing that deep learning models demonstrate high detection accuracy when trained and tested on the same Landsat-8 dataset (intra-dataset validation), outperforming operational algorithms. However, the performance of deep learning models is similar to operational threshold-based ones when tested on different Landsat-8 datasets (inter-dataset validation) or datasets from a different sensor with similar radiometric characteristics such as Sentinel-2 (cross-sensor validation).
Article
Environmental Sciences
Maciej Ziaja et al.
Summary: This paper introduces an end-to-end benchmarking approach for quantifying the abilities of deep learning algorithms in virtually any kind of on-board space applications. The experimental validation demonstrates that different deep learning techniques can be effectively benchmarked using standardized approach, delivering quantifiable performance measures and high configurability, which is crucial in delivering ready-to-use on-board artificial intelligence in emerging space applications.
Article
Multidisciplinary Sciences
Gonzalo Mateo-Garcia et al.
Summary: Spaceborne Earth observation technology provides valuable information for flood response; large constellations of small satellites can reduce revisit time in disaster areas; onboard processing helps reduce data transmission, with PhiSat-1 mission demonstrating hardware support for this approach.
SCIENTIFIC REPORTS
(2021)
Article
Multidisciplinary Sciences
B. Tellman et al.
Summary: Satellite imagery from 2000 to 2018 indicates that the proportion of global population exposed to floods is increasing, highlighting the importance of investing in flood adaptation strategies to reduce loss of life and livelihood. Changes in where and how floods occur, as well as the population exposed, are being influenced by urbanization, flood mitigation infrastructure, and settlements in floodplains. The high spatial and temporal resolution of satellite observations will enhance understanding of flood changes and adaptation methods.
Article
Computer Science, Artificial Intelligence
Yu Wang et al.
Summary: A novel method for ship detection using satellite images is proposed in this study, which involves preprocessing, feature extraction, and ship identification using machine learning classifiers. The proposed method is cross validated using Google Earth data and evaluated based on recall and precision values.
JOURNAL OF AMBIENT INTELLIGENCE AND SMART ENVIRONMENTS
(2021)
Article
Engineering, Electrical & Electronic
Gonzalo Mateo-Garcia et al.
Summary: The study introduces a domain adaptation transformation to reduce statistical differences between images from two satellite sensors in order to enhance the performance of transfer learning models.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Geography, Physical
Gonzalo Mateo-Garcia et al.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2020)
Article
Geochemistry & Geophysics
Onur Tasar et al.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2020)
Article
Environmental Sciences
Gianluca Giuffrida et al.
Article
Environmental Sciences
Edoardo Nemni et al.
Article
Environmental Sciences
Ruoyu Yang et al.
REMOTE SENSING APPLICATIONS-SOCIETY AND ENVIRONMENT
(2020)
Article
Multidisciplinary Sciences
Jean-Francois Pekel et al.
Article
Geochemistry & Geophysics
Devis Tuia et al.
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE
(2016)
Article
Remote Sensing
Hanqiu Xu
INTERNATIONAL JOURNAL OF REMOTE SENSING
(2006)