Initial Results of Site-Specific Assessment of Cereal Leaf Beetle (Oulema melanopus L.) Damage Using RGB Images by UAV.
Saved in:
| Title: | Initial Results of Site-Specific Assessment of Cereal Leaf Beetle (Oulema melanopus L.) Damage Using RGB Images by UAV. |
|---|---|
| Authors: | Sári-Barnácz, Fruzsina Enikő1 (AUTHOR), Kiss, Jozsef1,2 (AUTHOR) jozsef.kiss@uni-mate.hu, Kerezsi, György1,2 (AUTHOR), Szeredi, András Zoltán1,2 (AUTHOR), Pálinkás, Zoltán1 (AUTHOR), Zalai, Mihály1 (AUTHOR) |
| Source: | Remote Sensing. Jan2026, Vol. 18 Issue 1, p58. 44p. |
| Subjects: | Integrated pest control, Agricultural pests, Remote-sensing images, Agricultural drones, Leaf area, Precision farming, Machine learning |
| Geographic Terms: | Hungary |
| Abstract: | Highlights: What are the main findings? Vegetation index deviations from the field median yield beneficial predictors for random forest models in agricultural applications. The VARI, GCC, GLI, and NGRDI indices contain complementary information for detecting Oulema melanopus damage using RGB imagery. What are the implications of the main findings? The approach contributes to Integrated Pest Management decisions within precision agriculture and aids in reducing pesticide use. The presented approach contributes to the development of an operational, automated O. melanopus damage detection method for field application. Cereal leaf beetle (CLB, Oulema melanopus L.) is an important pest that damages cereals. Insecticide use against CLB could be reduced with targeted treatments. Our aims were to develop a methodology to map CLB damage on cereal fields using remote sensing. We investigated the suitability of four vegetation indices (VIs: the Visible Atmospherically Resistance Index (VARI), the Green Chromatic Coordinate (GCC), the Green Leaf Index (GLI), and the Normalized Green–Red Difference Index (NGRDI)) derived from RGB images (drone (UAV) imagery). Study sites were located in different regions of Hungary in 2024. Images were taken at different phenological stages of cereals. Suitability of VIs was analyzed with ANOVA and MANOVA. Machine learning models were developed to classify damaged field sections with random forest (RF) and Light Gradient Boosting Machine (LightGBM) algorithms. Results show that VARI, GCC, GLI, and NGRDI contain complementary features for early detection of CLB damage. Difference in sample points' VI from field median is advantageous for the LGBM algorithm (F1damaged = 0.64–0.72), while the best RF models were obtained with more features (F1damaged = 0.66). Random test data splits had optimistic results (overall accuracy: RF = 0.63–0.80, LightGBM = 0.63–0.79) compared to spatially controlled test splits (overall accuracy: RF = 0.53–0.70, LightGBM = 0.53–0.62). [ABSTRACT FROM AUTHOR] |
| Copyright of Remote Sensing is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | Highlights: What are the main findings? Vegetation index deviations from the field median yield beneficial predictors for random forest models in agricultural applications. The VARI, GCC, GLI, and NGRDI indices contain complementary information for detecting Oulema melanopus damage using RGB imagery. What are the implications of the main findings? The approach contributes to Integrated Pest Management decisions within precision agriculture and aids in reducing pesticide use. The presented approach contributes to the development of an operational, automated O. melanopus damage detection method for field application. Cereal leaf beetle (CLB, Oulema melanopus L.) is an important pest that damages cereals. Insecticide use against CLB could be reduced with targeted treatments. Our aims were to develop a methodology to map CLB damage on cereal fields using remote sensing. We investigated the suitability of four vegetation indices (VIs: the Visible Atmospherically Resistance Index (VARI), the Green Chromatic Coordinate (GCC), the Green Leaf Index (GLI), and the Normalized Green–Red Difference Index (NGRDI)) derived from RGB images (drone (UAV) imagery). Study sites were located in different regions of Hungary in 2024. Images were taken at different phenological stages of cereals. Suitability of VIs was analyzed with ANOVA and MANOVA. Machine learning models were developed to classify damaged field sections with random forest (RF) and Light Gradient Boosting Machine (LightGBM) algorithms. Results show that VARI, GCC, GLI, and NGRDI contain complementary features for early detection of CLB damage. Difference in sample points' VI from field median is advantageous for the LGBM algorithm (F1damaged = 0.64–0.72), while the best RF models were obtained with more features (F1damaged = 0.66). Random test data splits had optimistic results (overall accuracy: RF = 0.63–0.80, LightGBM = 0.63–0.79) compared to spatially controlled test splits (overall accuracy: RF = 0.53–0.70, LightGBM = 0.53–0.62). [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 20724292 |
| DOI: | 10.3390/rs18010058 |
