Discrimination of fungal disease infestation in oil-palm canopy hyperspectral reflectance data
This study focuses on the calibration of a statistical model of discrimination between different stages of a fungal disease attack on oil palm, based on field hyperspectral measurements at the canopy scale. Combinations of preprocessing, partial least square regression and factorial discriminant analysis are tested on a hundred of samples to prove the efficiency of canopy reflectance to provide information about the plant sanitary status. A robust algorithm is thus derived, allowing classifying oil palm in a 4-level typology, based on disease severity levels from the sane to the critically sick tree with a global performance of more than 92%. Applications and further improvements of this experiment are discussed.
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | conference_item biblioteca |
| Language: | eng |
| Published: |
s.n.
|
| Subjects: | H20 - Maladies des plantes, U10 - Informatique, mathématiques et statistiques, U30 - Méthodes de recherche, Elaeis guineensis, Ganoderma, http://aims.fao.org/aos/agrovoc/c_2509, http://aims.fao.org/aos/agrovoc/c_15973, http://aims.fao.org/aos/agrovoc/c_7518, |
| Online Access: | http://agritrop.cirad.fr/553401/ http://agritrop.cirad.fr/553401/1/document_553401.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study focuses on the calibration of a statistical model of discrimination between different stages of a fungal disease attack on oil palm, based on field hyperspectral measurements at the canopy scale. Combinations of preprocessing, partial least square regression and factorial discriminant analysis are tested on a hundred of samples to prove the efficiency of canopy reflectance to provide information about the plant sanitary status. A robust algorithm is thus derived, allowing classifying oil palm in a 4-level typology, based on disease severity levels from the sane to the critically sick tree with a global performance of more than 92%. Applications and further improvements of this experiment are discussed. |
|---|