We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
A COMPARATIVE STUDY OF SATELLITE IMAGE RESOLUTIONS FOR DETECTING PEST DAMAGE IN SUNFLOWER FIELDS.
- Authors
Ercan, B. S.; Maden, B.; Kara, S.; Sunar, F.; Aysal, T.; Ozkaya, N.; Saglam, O.
- Abstract
The diversity of sensors in remote sensing allows for faster and easier detection of changes and issues across different scales, in contrast to conventional ground-based systems. One of the most important distinguishing features among these sensors is their varying resolutions, contributing to the versatility of remote sensing technologies across diverse environmental applications. In this study, the effectiveness of PlanetScope and Sentinel-2 satellite images with different image resolutions in detecting damage caused by a harmful insect (<em>beet webworm moth - Loxostege sticticalis</em>) in sunflower fields in Lüleburgaz district of Kırklareli in the Trace region was evaluated. Damage rates in sunflower fields were analyzed using various spectral indices (Enhanced Vegetation Index and Chlorophyll Index Green) and spectral transformation (Tasseled Cap Greenness) in conjunction with in situ data. Based on the spectral analysis, the satellite image dated 26 July, which showed the most severe damage, was used in the damage assessment analysis. The damaged areas were compared by classifying both satellite images with the Random Forest algorithm. According to the results of the classification accuracy assessment, PlanetScope satellite imagery showed the highest accuracy, with 90% overall accuracy and 84% Kappa statistics, making it a more suitable sensor choice for agricultural applications.
- Subjects
REMOTE-sensing images; RANDOM forest algorithms; REMOTE sensing; SUNFLOWERS
- Publication
International Archives of the Photogrammetry, Remote Sensing & Spatial Information Sciences, 2024, Vol 48, Issue 4/W9, p133
- ISSN
1682-1750
- Publication type
Article
- DOI
10.5194/isprs-archives-XLVIII-4-W9-2024-133-2024