Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks

Remote Sensing, Год журнала: 2024, Номер 16(5), С. 823 - 823

Опубликована: Фев. 27, 2024

Leveraging mid-resolution satellite images such as Landsat 8 for accurate farmland segmentation and land change monitoring is crucial agricultural management, yet hindered by the scarcity of labelled data training supervised deep learning pipelines. The particular focus this study on addressing images. This paper introduces several contributions, including a systematic image augmentation approach that aims to maintain population consistency during model training, thus mitigating performance degradation. To alleviate labour-intensive task pixel-wise labelling, we present novel application modified conditional generative adversarial network (CGAN) generate artificial corresponding farm labels. Additionally, scrutinize role spectral bands in compare two prominent semantic models, U-Net DeepLabV3+, with diverse backbone structures. Our empirical findings demonstrate augmenting dataset up 22.85% samples significantly enhances performance. Notably, model, employing standard convolution, outperforms DeepLabV3+ models atrous achieving accuracy 86.92% test data.

Язык: Английский

---

Agricultural sustainability assessment and national policy-making using an axiomatic mathematical model

Environmental and Sustainability Indicators, Год журнала: 2024, Номер 22, С. 100401 - 100401

Опубликована: Май 1, 2024

Agriculture is a basic activity of human survival and its sustainability the utmost importance. In this paper, we present novel mathematical model for assessment agricultural sustainability, based on certain intuitive fundamental postulates. The resulting framework generalizes most existing models. We then rank 148 countries according to their using data from period 1995 2022 pinpoint those aspects with highest potential improving sensitivity analysis. Our findings demonstrate that farming mostly unsustainable worldwide. crucial factors are connected bad water use, destruction biodiversity, prevalence conventional chemical farming. As expected, poor make bottom list. Surprisingly, several advanced such as Belgium, Japan, Netherlands Norway also occupy list because very high emissions performance concerning land biodiversity. Finally, score worldwide order only 70% median 50%, which exhibit rather unsatisfactory state affairs. Such ought guide immediate action agriculture more sustainable.

Язык: Английский

---

Assessment of the impact of agriculture on the regional socio-economic development

E3S Web of Conferences, Год журнала: 2024, Номер 548, С. 01003 - 01003

Опубликована: Янв. 1, 2024

This article analyzed the impact of agriculture on regional economy, population employment and export potential. Studies show that volume agricultural production its has a significant region's gross product, income. Also, study shows coefficient elasticity products is high, development growth economy. In addition, emphasized importance developing agriculture-related industries services for sustainable growth. article, regression correlation analyzes were used to determine relationship between other economic indicators.

Язык: Английский

---

Investigating the Spatial Distribution and Influencing Factors of Non-Grain Production of Farmland in South China Based on MaxEnt Modeling and Multisource Earth Observation Data

Foods, Год журнала: 2024, Номер 13(21), С. 3385 - 3385

Опубликована: Окт. 24, 2024

Excessive non-grain production of farmland (NGPF) seriously affects food security and hinders progress toward Sustainable Development Goal 2 (Zero Hunger). Understanding the spatial distribution influencing factors NGPF is essential for agricultural management. However, previous studies on identification have mainly relied high-cost methods (e.g., visual interpretation). Furthermore, common machine learning techniques difficulty in accurately identifying based solely spectral information, as not merely a natural phenomenon. Accurately at grid scale elucidating its emerged critical scientific challenges current literature. Therefore, aims this study are to develop grid-scale method that integrates multisource remote sensing data enhance precision provide more comprehensive understanding factors. To overcome these challenges, we combined images, natural/anthropogenic factors, maximum entropy model reveal scale. This combination can detailed information quantify integrated influences multiple from microscale perspective. In case Foshan, China, area under receiver operating characteristic curve 0.786, with results differing by only 1.74% statistical yearbook results, demonstrating reliability method. Additionally, total error our result lower than using information. Our enhances resolution effectively detects small fragmented farmlands. We identified elevation, farming radius, population density dominant affecting NGPF. These offer targeted strategies mitigate excessive The advantage lies independence negative samples. feature applicability other cases, particularly regions lacking high-resolution grain crop-related data.

Язык: Английский

---

A novel evaluation of farmland soil environmental risk integrating heavy metal(loid) pollution risk and industrial risk potential

Stochastic Environmental Research and Risk Assessment, Год журнала: 2025, Номер unknown

Опубликована: Май 27, 2025

Язык: Английский

---