Burned Olive Trees Identification with a Deep Learning Approach in Unmanned Aerial Vehicle Images DOI Creative Commons
Christos Vasilakos, Vassilios S. Verykios

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(23), P. 4531 - 4531

Published: Dec. 3, 2024

Olive tree orchards are suffering from wildfires in many Mediterranean countries. Following a wildfire event, identifying damaged olive trees is crucial for developing effective management and restoration strategies, while rapid damage assessment can support potential compensation producers. Moreover, the implementation of real-time health monitoring groves allows producers to carry out targeted interventions, reducing production losses preserving crop health. This research examines use deep learning methodologies true-color images Unmanned Aerial Vehicles (UAV) detect trees, including withering desiccation branches leaf scorching. More specifically, object detection image classification computer vision techniques area applied compared. In approach, algorithm aims localize identify burned/dry unburned/healthy classifier categorizes an showing as or unburned/healthy. Training data included true color UAV by fire obtained multiple cameras flight heights, resulting various resolutions. For detection, Residual Neural Network was used backbone approach with Single-Shot Detector. application, two approaches were evaluated. first new shallow network developed, second transfer pre-trained networks applied. According results, managed healthy average accuracy 74%, drying, 69%. However, optimal identified (healthy unhealthy) that user did not during collection. application convolutional neural achieved significantly better results F1-score above 0.94, either training applying learning. conclusion, performed than detection.

Language: Английский

Burned Areas Mapping Using Sentinel-2 Data and a Rao’s Q Index-Based Change Detection Approach: A Case Study in Three Mediterranean Islands’ Wildfires (2019–2022) DOI Creative Commons
Rafaela Tiengo, S. Merino de Miguel, Jéssica Uchôa

et al.

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(5), P. 830 - 830

Published: Feb. 27, 2025

This study explores the application of remote sensing-based land cover change detection techniques to identify and map areas affected by three distinct wildfire events that occurred in Mediterranean islands between 2019 2022, namely Sardinia (2019, Italy), Thassos (2022, Greece), Pantelleria Italy). Applying Rao’s Q Index-based approach Sentinel-2 spectral data derived indices, we evaluate their effectiveness accuracy identifying mapping burned wildfires. Our methodological implies processing analysis pre- post-fire imagery extract relevant indices such as Normalized Burn Ratio (NBR), Mid-infrared Index (MIRBI), Difference Vegetation (NDVI), Burned area for (BAIS2) then use (the classic approach) or combine them (multidimensional detect using a technique. The Copernicus Emergency Management System (CEMS) were used assess validate all results. lowest overall (OA) classical mode was 52%, BAIS2 index, while multidimensional mode, it 73%, combining NBR NDVI. highest result reached 72% with MIRBI 96%, NBR. combination consistently achieved across areas, demonstrating its improving classification regardless characteristics.

Language: Английский

Citations

1
Satellite‐Aided Disaster Response DOI Creative Commons
J. Rolla, Aditya Khuller, Karen An

et al.

AGU Advances, Journal Year: 2025, Volume and Issue: 6(1)

Published: Feb. 1, 2025

Abstract The increasing frequency and severity of natural disasters, driven by climate change anthropogenic activities, pose unprecedented challenges to emergency response agencies worldwide. Satellite remote sensing has become a critical tool for providing timely accurate data aid in disaster preparedness, response, recovery. This Commentary explores the role satellite managing climate‐driven highlighting use technologies such as Synthetic Aperture Radar (SAR) creating damage proxy maps. These maps are instrumental assessing impacts guiding efforts, demonstrated 2023 Wildfires Hawaii. Despite promise these tools, remain, including need rapid processing, automation pipelines, robust international collaborations. future missions composing Earth System Observatory, upcoming NASA‐ISRO SAR mission, represents significant advancement with its global coverage frequent, detailed measurements. study emphasizes importance continued investment advanced cooperation enhance capabilities, ultimately building more resilient community.

Language: Английский

Citations

0
Spatial Agreement of Burned Area Products Derived from Very High to Coarse-Resolution Satellite Imagery in African Biomes DOI Creative Commons
Daniela Stroppiana, Matteo Sali, Pietro Alessandro Brivio

et al.

Fire, Journal Year: 2025, Volume and Issue: 8(4), P. 126 - 126

Published: March 26, 2025

Satellite data provide the spatial distributions of burned areas worldwide; assessing their accuracy and comparing area estimates from different products is relevant to gain insights into reliability sources error. We compared BA maps derived multispectral satellite with resolutions, ranging Planet (3 m) Sentinel-2 (S2, 10–20 m), Sentinel-3 (S3, 300 MODIS (250–500 over selected African sites for year 2019. S2 images were processed derive a supervised Random Forest algorithm used assess agreement FireCCISFD20, FireCCI51, FireCCIS311, MCD64A1 by computing omission commission errors, Dice Coefficient, Relative bias. The based on showed greatest very high-resolution (overall Coefficient was found be greater than 80%). coarse-resolution lower reference perimeters. Among coarse resolution products, FireCCIS311 outperform others. influential accuracy, error (RelB < 0) coarser products. patterns burns vegetation type significant in mapping detection Sahelian savannas more accurate. This study provides variability high- imagery.

Language: Английский

Citations

0
Burned Olive Trees Identification with a Deep Learning Approach in Unmanned Aerial Vehicle Images DOI Creative Commons
Christos Vasilakos, Vassilios S. Verykios

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(23), P. 4531 - 4531

Published: Dec. 3, 2024

Olive tree orchards are suffering from wildfires in many Mediterranean countries. Following a wildfire event, identifying damaged olive trees is crucial for developing effective management and restoration strategies, while rapid damage assessment can support potential compensation producers. Moreover, the implementation of real-time health monitoring groves allows producers to carry out targeted interventions, reducing production losses preserving crop health. This research examines use deep learning methodologies true-color images Unmanned Aerial Vehicles (UAV) detect trees, including withering desiccation branches leaf scorching. More specifically, object detection image classification computer vision techniques area applied compared. In approach, algorithm aims localize identify burned/dry unburned/healthy classifier categorizes an showing as or unburned/healthy. Training data included true color UAV by fire obtained multiple cameras flight heights, resulting various resolutions. For detection, Residual Neural Network was used backbone approach with Single-Shot Detector. application, two approaches were evaluated. first new shallow network developed, second transfer pre-trained networks applied. According results, managed healthy average accuracy 74%, drying, 69%. However, optimal identified (healthy unhealthy) that user did not during collection. application convolutional neural achieved significantly better results F1-score above 0.94, either training applying learning. conclusion, performed than detection.

Language: Английский

Citations

0