Articles
Evaluation of UAV and Sentinel 2 images to estimate condition of hop (Humulus lupulus L.) plants
Article number
1328_13
Pages
95 – 102
Language
English
Abstract
Hop garden structure and crop condition can be monitored by many in situ methods but remote sensing offers possibilities of non-destructive monitoring.
An organic hop garden with the cultivars atecký červeňák and Premiant, and a conventional hop garden with the cultivars Agnus, Premiant and Sládek were selected for the experiment.
In this study, unmanned aerial vehicle (UAV, eBee X senseFly UAV) equipped with Red Green Blue (RGB), Sensor Optimized for Drone Application (S.O.D.A.), and Sequoia cameras as well as Sentinel 2 satellite images were used for monitoring of hop plants throughout the whole vegetation season.
UAV images were pre-processed in Pix4Dmapper software and orthophoto mosaics were derived.
RGB vegetation indices and normalized difference vegetation index (NDVI) were calculated for each cultivar from UAV and satellite images.
Resulted binary model of portion with green leaves and soil was extracted on the base of threshold value and served for detection of plants area.
NDVI spectral index was calculated in order to describe plant condition during the growing season.
The results showed significant differences in NDVI values among the conventional and organic hop gardens and among the cultivars as well and showed applicability of UAV and satellite images.
Green area of the plants derived from the UAV images was influenced by design and flight parameters of this unmanned aerial system and crop structure.
An organic hop garden with the cultivars atecký červeňák and Premiant, and a conventional hop garden with the cultivars Agnus, Premiant and Sládek were selected for the experiment.
In this study, unmanned aerial vehicle (UAV, eBee X senseFly UAV) equipped with Red Green Blue (RGB), Sensor Optimized for Drone Application (S.O.D.A.), and Sequoia cameras as well as Sentinel 2 satellite images were used for monitoring of hop plants throughout the whole vegetation season.
UAV images were pre-processed in Pix4Dmapper software and orthophoto mosaics were derived.
RGB vegetation indices and normalized difference vegetation index (NDVI) were calculated for each cultivar from UAV and satellite images.
Resulted binary model of portion with green leaves and soil was extracted on the base of threshold value and served for detection of plants area.
NDVI spectral index was calculated in order to describe plant condition during the growing season.
The results showed significant differences in NDVI values among the conventional and organic hop gardens and among the cultivars as well and showed applicability of UAV and satellite images.
Green area of the plants derived from the UAV images was influenced by design and flight parameters of this unmanned aerial system and crop structure.
Publication
Authors
J. Kumhálová, J. Chyba, K. Krofta, K. Starý, V. Brant
Keywords
unmanned aerial system, satellite images, spectral indices, cultivars, cultivation
Online Articles (21)