A Machine Learning Approach to Detecting Pine Wilt Disease Using Airborne Spectral Imagery
| Autor: | Vasco Mantas, Elsa Baltazar, N. Lewyckyj, Marian-Daniel Iordache, Klaas Pauly |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
010504 meteorology & atmospheric sciences
Computer science multispectral Multispectral image 0211 other engineering and technologies Bursaphelenchus xylophilus 02 engineering and technology 01 natural sciences remote sensing lcsh:Science 021101 geological & geomatics engineering 0105 earth and related environmental sciences Wilt disease biology Sampling (statistics) Hyperspectral imaging Pine Wilt Disease machine learning classification hyperspectral early detection remotely piloted aircraft systems 15. Life on land biology.organism_classification Random forest Statistical classification General Earth and Planetary Sciences Pinus pinaster lcsh:Q Cartography |
| Zdroj: | Remote Sensing Remote Sensing, Vol 12, Iss 2280, p 2280 (2020) Remote Sensing; Volume 12; Issue 14; Pages: 2280 |
| Popis: | Pine Wilt Disease is one of the most destructive pests affecting coniferous forests. After being infected by the harmfulBursaphelenchus xylophilusnematode, most trees die within one year. The complex spreading pattern of the disease and the tedious hard labor process of diagnosis involving field wood sampling followed by laboratory analysis call for alternative methods to detect and manage the infected areas. Remote sensing comes naturally into play owing to the possibility of covering relatively large areas and the ability to discriminate healthy from sick trees based on spectral characteristics. This paper presents the development of machine learning classification algorithms for the detection of Pine Wilt Disease inPinus pinaster, performed in the framework of the European Commission’s Horizon 2020 project “Operational Forest Monitoring using Copernicus and UAV Hyperspectral Data” (FOCUS) in two provinces of central Portugal. Five flight campaigns have been carried out in two consecutive years in order to capture a multitemporal variation of disease distribution. Classification algorithms based on a Random Forest approach were separately designed for the acquired very-high-resolution multispectral and hyperspectral data, respectively. Both algorithms achieved overall accuracies higher than 0.91 in test data. Furthermore, our study shows that the early detection of decaying trees is feasible, even before symptoms are visible in the field. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|