GiESCO 2019 banner
IVES 9 IVES Conference Series 9 GiESCO 9 Non-linear unmixing as an innovative tool to detect vine diseases in UAVs, airborned and satellite images: preliminary results

Non-linear unmixing as an innovative tool to detect vine diseases in UAVs, airborned and satellite images: preliminary results

Abstract

Context and purpose of the study – Vine diseases have a strong impact on vineyards sustainability, which in turns leads to strong economic consequences. Among those diseases, Flavescence dorée spreads quickly and is incurable, which led in France to the setup of a mandatory pest control implying the systematic use of pesticides and the prospection and uprooting of every infected plants. Remote sensing could be a very powerful tool to optimize prospection as it allows to produce quickly accurate maps over large areas. Recent studies have shown that high spatial resolution (10cm/pixel) multispectral images acquired from UAVs allow to map Flavescence dorée in vineyards using leaves discolorations [e.g. Albetis et al., Remote Sensing, 2017]. Nevertheless, confusion and misdetections still exist, especially with other diseases showing similar leaves discolorations and with mixtures of different materials occurring within one pixel. Mixture effects are also crucial when dealing with satellite images where spatial resolution is much lower (≥10m/pixel). This study aims at improving the detection of vine diseases in UAVs, airborned and satellite images using an innovative tool that identifies the spectral signatures of every elementary materials (e.g. healthy and sick leaves) and their relative contribution at a subpixel level.

Material and methods – We use three distinct datasets acquired in 2016 over the same vineyard located in the Southwest of France (AOC Gaillac): a multispectral image acquired with MicaSense sensor onboard an UAV (5 bands, 10cm/pixel), a Sentinel-2 multispectral image (12 bands, 10m/pixel) and an airborned hyperspectral image (256 bands, 1m/pixel). Ground truth for validation is available through exhaustive centimetric locations of every sick vines for several plots in the studied area. On the methodological perspective, we use an innovative method that performs an unsupervised unmixing jointly with anomalydetection capacities and has a global linear complexity [Nakhostin et al., TGRS, 2016]. Nonlinearities are handled by decomposing the data on an overcomplete set of spectra, combined with a specific sparse projection, which guarantees the interpretability of the analysis.

Results – This paper reports preliminary results obtained with the unmixing algorithm ran over one selected plot available in the dataset. Initial results show the algorithm can detect and separate multiple sources within the plot. Analysis of retrieved endmembers shows a good correlation with the components that can be found in the field, especially with the evidence of healthy and sick leaves’ signatures. Nevertheless, initial mapping still shows some discrepancies with ground truth and further work needs to be done to fine tune the model parameters.

DOI:

Publication date: September 27, 2023

Issue: GiESCO 2019

Type: Poster

Authors

Harold CLENET1,2*, Sina NAKHOSTIN3, Eve LAROCHE-PINEL1,2,4, Sylvie DUTHOIT4

1 Ecole d’Ingénieurs de PURPAN, Toulouse INP, 75 voie du TOEC, 31076 Toulouse, France
2 UMR 1201 DYNAFOR, INRA-Toulouse INP, Chemin de Borde-Rouge, 31326 Castanet-Tolosan, France
3 Ecole et Observatoire des Sciences de la Terre – EOST, 67084 Strasbourg, France
4 TerraNIS, 12 Avenue de l’Europe, 31520 Ramonville Saint-Agne, France

Contact the author

Keywords

vine diseases, remote sensing, image processing, non-linear unmixing, satellite imagery, UAVs

Tags

GiESCO | GiESCO 2019 | IVES Conference Series

Citation

Related articles…

Effect of simulated shipping conditions on colour and SO2 evolution in soave wines

The shelf life of food is defined as the period in which the product will remain safe, is certain to retain desired sensory, chemical, physical, and microbiological characteristics

Evaluation of Valdadige DOC “Terra dei Forti” vineyards by zoning approach

La conoscenza dell’interazione genotipo x ambiente e pertanto della caratterizzazione territoriale è di prioritaria importanza nella valutazione dei siti. Grazie alla combinazione di dati GIS

Elevational range shifts of mountain vineyards: Recent dynamics in response to a warming climate

Increasing temperatures worldwide are expected to cause a change in spatial distribution of plant species along elevational gradients and there are already observable shifts to higher elevations as a consequence of climate change for many species. Not only naturally growing plants, but also agricultural cultivations are subject to the effects of climate change, as the type of cultivation and the economic viability depends largely on the prevailing climatic conditions. A shift to higher elevations therefore represents a viable adaptation strategy to climate change, as higher elevations are characterized by lower temperatures. This is especially important in the case of viticulture because a certain wine-style can only be achieved under very specific climatic conditions. Although there are several studies investigating climatic suitability within winegrowing regions or longitudinal shifts of winegrowing areas, little is known about how fast vineyards move to higher elevations, which may represent a viable strategy for winegrowers to maintain growing conditions and thus wine-style, despite the effects of climate change. We therefore investigated the change in the spatial distribution of vineyards along an elevational gradient over the past 20 years in the mountainous wine-growing region of Alto Adige (Italy). A dataset containing information about location and planting year of more than 26000 vineyard parcels and 30 varieties was used to perform this analysis. Preliminary results suggest that there has been a shift to higher elevations for vineyards in general (from formerly 700m to currently 850 m a.s.l., with extreme sites reaching 1200 m a.s.l.), but also that this development has not been uniform across different varieties and products (i.e. vitis vinifera vs hybrid varieties and still vssparkling wines). This is important for climate change adaptation as well as for rural development. Mountain areas, especially at mid to high elevations, are often characterized by severe land abandonment which can be avoided to some degree if economically viable and sustainable land management strategies are available.

Identifying wild Vitis riparia Michx clones as a source of rootstock to mitigate vigour and acclimation/deacclimation cycles of the scion

Grapevine rootstocks have traditionally been chosen in order to manage scion vigour, soil pests and soil conditions. Riparia Gloire de Montpellier (RGM) has been in use since the turn of the 19th century, over 100 years and still a remarkably stable source of phylloxera (Daktulosphaeria vitifoliae Fitch) resistance. The original source material was probably collected near the Missouri/Mississippi river confluence, a mid-continental but more southerly location in the United States. It has been hypothesized that more northerly selections of V. riparia Michx might improve both fall acclimation rate and depth of the scion, thus mitigating late fall frost and midwinter freeze damage.

Detection of spider mite using artificial intelligence in digital viticulture

Aim: Pests have a high impact on yield and grape quality in viticulture. An objective and rapid detection of pests under field conditions is needed. New sensing technologies and artificial intelligence could be used for pests detection in digital viticulture. The aim of this work was to apply computer vision and deep learning techniques for automatic detection of spider mite symptoms in grapevine under field conditions.