Federated LeViT-ResUNet for Scalable and Privacy-Preserving Agricultural Monitoring Using Drone and Internet of Things Data

Agronomy, Journal Year: 2025, Volume and Issue: 15(4), P. 928 - 928

Published: April 10, 2025

Precision agriculture is necessary for dealing with problems like pest outbreaks, a lack of water, and declining crop health. Manual inspections broad-spectrum pesticide application are inefficient, time-consuming, dangerous. New drone photography IoT sensors offer quick, high-resolution, multimodal agricultural data collecting. Regional diversity, heterogeneity, privacy make it hard to conclude these data. This study proposes lightweight, hybrid deep learning architecture called federated LeViT-ResUNet that combines the spatial efficiency LeViT transformers ResUNet’s exact pixel-level segmentation address issues. The system uses multispectral footage sensor identify real-time insect hotspots, health, yield prediction. dynamic relevance sparsity-based feature selector (DRS-FS) improves ranking reduces redundancy. Spectral normalization, spatial–temporal alignment, dimensionality reduction provide reliable input representation. Unlike centralized models, our platform trains over-dispersed client datasets using preserve capture regional trends. A huge, open-access dataset from varied environmental circumstances was used simulation experiments. suggested approach on conventional models ResNet, DenseNet, vision transformer 98.9% classification accuracy 99.3% AUC. scalable sustainable privacy-preserving precision because its high generalization, low latency, communication efficiency. lays groundwork real-time, intelligent monitoring systems in diverse, resource-constrained farming situations.

Language: Английский

---

A Systematic Review on the Advancements in Remote Sensing and Proximity Tools for Grapevine Disease Detection

Sensors, Journal Year: 2024, Volume and Issue: 24(24), P. 8172 - 8172

Published: Dec. 21, 2024

Grapevines (Vitis vinifera L.) are one of the most economically relevant crops worldwide, yet they highly vulnerable to various diseases, causing substantial economic losses for winegrowers. This systematic review evaluates application remote sensing and proximal tools vineyard disease detection, addressing current capabilities, gaps, future directions in sensor-based field monitoring grapevine diseases. The covers 104 studies published between 2008 October 2024, identified through searches Scopus Web Science, conducted on 25 January updated 10 2024. included focused exclusively detection while excluded were not related did use or sensing, conditions. studied diseases include downy mildew, powdery Flavescence dorée, esca complex, rots, viral main sensors RGB, multispectral, hyperspectral sensors, spectroscopy. A trend recent research is integration artificial intelligence techniques, such as machine learning deep learning, improve accuracy. results demonstrate progress monitoring, with concentrating specific sensor platforms, methodological improvements. Future should focus standardizing methodologies, integrating multi-sensor data, validating approaches across diverse contexts commercial applicability sustainability, both environmental challenges.

Language: Английский

---

The utility of Radiative Transfer Models (RTM) on remotely sensed data in retrieving biophysical and biochemical properties of terrestrial biomes: A Systematic Review

Advances in Space Research, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 1, 2025

Language: Английский

---

Improving maize water stress diagnosis accuracy by integrating multimodal UAVs data and leaf area index inversion model

Agricultural Water Management, Journal Year: 2025, Volume and Issue: 312, P. 109407 - 109407

Published: March 9, 2025

Language: Английский

---

Improving chili pepper LAI prediction with TPE-2BVIs and UAV hyperspectral imagery

Computers and Electronics in Agriculture, Journal Year: 2025, Volume and Issue: 235, P. 110368 - 110368

Published: April 4, 2025

Language: Английский

---

A Comprehensive Review of Crop Chlorophyll Mapping Using Remote Sensing Approaches: Achievements, Limitations, and Future Perspectives

Sensors, Journal Year: 2025, Volume and Issue: 25(8), P. 2345 - 2345

Published: April 8, 2025

Chlorophyll absorbs light energy and converts it into chemical energy, making a crucial biochemical parameter for monitoring vegetation health, detecting environmental stress, predicting physiological states. Accurate rapid estimation of canopy chlorophyll content is assessing dynamics, ecological changes, growth patterns. Remote sensing technology has become an indispensable tool since 2015, with more than 50 research papers published annually, contributing to substantial body case studies. This review discusses remote technologies currently used estimating content, focusing on four key aspects: the acquisition reference datasets, identification optimal spectral variables, selection models, analysis application scenarios. The results indicate that bands in visible red-edge regions (e.g., 530 nm, 670 705 nm) provide high prediction accuracy. Machine learning methods, such as random forest support vector regression, exhibit excellent performance, determination coefficients (R2) typically exceeding 0.9, although overfitting remains issue. Although radiative transfer models are slightly less accurate (R2 = 0.6–0.8), they greater interpretability. Hybrid integrating machine show strong potential balance accuracy generalizability. Future should improve model generalizability different types conditions integrate multi-source data spatial temporal resolution. Combining physical processing artificial intelligence, can scalability, cost-effectiveness, real-time capabilities.

Language: Английский

---

Estimating stomatal conductance of citrus orchard based on UAV multi-modal information in Southwest China

Agricultural Water Management, Journal Year: 2024, Volume and Issue: 307, P. 109253 - 109253

Published: Dec. 19, 2024

Language: Английский

---