Clean Collector Algorithm for Satellite Image Pre-Processing of SAR-to-EO Translation DOI Open Access
Min-Woo Kim, Sekil Park, Jingi Ju

et al.

Electronics, Journal Year: 2024, Volume and Issue: 13(22), P. 4529 - 4529

Published: Nov. 18, 2024

In applications such as environmental monitoring, algorithms and deep learning-based methods using synthetic aperture radar (SAR) electro-optical (EO) data have been proposed with promising results. These results achieved already cleaned datasets for training data. However, in real-world collection, are often collected regardless of noises (clouds, night, missing data, etc.). Without cleaning the these noises, trained model has a critical problem poor performance. To address issues, we propose Clean Collector Algorithm (CCA). First, use pixel-based approach to clean QA60 mask outliers. Secondly, remove night-time that can act noise process. Finally, feature-based refinement method cloud images FID. We demonstrate its effectiveness by winning first place SAR-to-EO translation track MultiEarth 2023 challenge. also highlight performance robustness CCA on other datasets, SEN12MS-CR-TS Scotland&India.

Language: Английский

Analyzing the effects of climate change and human activities on streamflow in a North China arid basin: a machine learning perspective considering model structural uncertainty DOI
Jinqiang Wang, Ling Zhou,

Chi Ma

et al.

Theoretical and Applied Climatology, Journal Year: 2025, Volume and Issue: 156(2)

Published: Jan. 18, 2025

Language: Английский

Citations

1
Rainwater harvesting technologies in arid plains of Argentina: small local strategies vs. large centralized projects DOI Creative Commons
Aldana Calderón Archina, Diego Escolar, Guillermo Heider

et al.

Frontiers in Environmental Science, Journal Year: 2024, Volume and Issue: 12

Published: Nov. 13, 2024

Access to water has been and remains one of humanity’s greatest challenges. Especially in arid plains exposed significant climatic fluctuations future global change trends. In the past present, local communities central-western Argentina (i.e., Guanacache Lagoons, Cuyo region) have developed multiple strategies manage supply problems. The aims this study are: i) characterize different harvesting technologies (pre-Hispanic modern) used, ii) compare small (bottom-up solutions) with large centralized projects (top-down solutions). On hand, we show transformations these over time, challenges faced by inhabitants context climate other analyze role state through hydraulic policies implemented provincial states last two centuries how impacted area. This review is based on a historical archaeological bibliography, recent publications about region, including articles our ethnographic fieldwork. Our results demonstrate valuable experience accumulated populations methods, particularly areas where groundwater deep saline, shows adaptability contexts increasing scarcity. We considered that indigenous knowledge can largely contribute sustainable management resources. might be useful for decision-makers managers drylands around world find equitable approach combines technical advances knowledge.

Language: Английский

Citations

0
Clean Collector Algorithm for Satellite Image Pre-Processing of SAR-to-EO Translation DOI Open Access
Min-Woo Kim, Sekil Park, Jingi Ju

et al.

Electronics, Journal Year: 2024, Volume and Issue: 13(22), P. 4529 - 4529

Published: Nov. 18, 2024

In applications such as environmental monitoring, algorithms and deep learning-based methods using synthetic aperture radar (SAR) electro-optical (EO) data have been proposed with promising results. These results achieved already cleaned datasets for training data. However, in real-world collection, are often collected regardless of noises (clouds, night, missing data, etc.). Without cleaning the these noises, trained model has a critical problem poor performance. To address issues, we propose Clean Collector Algorithm (CCA). First, use pixel-based approach to clean QA60 mask outliers. Secondly, remove night-time that can act noise process. Finally, feature-based refinement method cloud images FID. We demonstrate its effectiveness by winning first place SAR-to-EO translation track MultiEarth 2023 challenge. also highlight performance robustness CCA on other datasets, SEN12MS-CR-TS Scotland&India.

Language: Английский

Citations

0