Indirect measurements of Mediterranean grassland vegetation characteristics using plenoptic camera and NIRS
The description of the grassland vegetation and its multiple functions such as forage production, biodiversity and nutritive values requires the collection of many types of data on the field. This collection is in general cumbersome and laborious (e.g. counting the plants by species in sampled quadrats) and, when destructive, it does not allow to monitor the dynamics of the vegetation during the growth season. As an alternative, indirect measurements can be done using optical tools. In this work, we tested the interest of combining plenoptic images and NIRS information to characterize different variables of Mediterranean grasslands. A Plenoptic camerais a multifocal camera that produces clear images on all the deepness of the images. We tested this protocol on two grassland mixtures for estimating the functional diversity of the mixtures. NIRS and plenoptic images were taken at the same time. The images were analyzed using Trainable Weka Segmentation ImageJ. The outputs of the images and NIRS analyses were afterwards correlated with the field measurements.
| Main Authors: | , , , , , , , |
|---|---|
| Format: | conference_item biblioteca |
| Language: | eng |
| Published: |
Wageningen Academic Publishers
|
| Subjects: | U30 - Méthodes de recherche, L02 - Alimentation animale, P01 - Conservation de la nature et ressources foncières, |
| Online Access: | http://agritrop.cirad.fr/585518/ http://agritrop.cirad.fr/585518/1/ID585518.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The description of the grassland vegetation and its multiple functions such as forage production, biodiversity and nutritive values requires the collection of many types of data on the field. This collection is in general cumbersome and laborious (e.g. counting the plants by species in sampled quadrats) and, when destructive, it does not allow to monitor the dynamics of the vegetation during the growth season. As an alternative, indirect measurements can be done using optical tools. In this work, we tested the interest of combining plenoptic images and NIRS information to characterize different variables of Mediterranean grasslands. A Plenoptic camerais a multifocal camera that produces clear images on all the deepness of the images. We tested this protocol on two grassland mixtures for estimating the functional diversity of the mixtures. NIRS and plenoptic images were taken at the same time. The images were analyzed using Trainable Weka Segmentation ImageJ. The outputs of the images and NIRS analyses were afterwards correlated with the field measurements. |
|---|