Terroir 2006 banner
IVES 9 IVES Conference Series 9 Vine field monitoring using high resolution remote sensing images: segmentation and characterization of rows of vines

Vine field monitoring using high resolution remote sensing images: segmentation and characterization of rows of vines

Abstract

A new framework for the segmentation and characterization of row crops on remote sensing images has been developed and validated for vineyard monitoring. This framework operates on any high-resolution remote sensing images since it is mainly based on geometric information. It aims at obtaining maps describing the variation of a vegetation index such as NDVI along each row of a parcel.
The framework consists in several steps. First, the segmentation step allows the delineation of the parcel under consideration. A region-growing algorithm, based on the textural properties of row crops, was developed for this purpose. Once the parcel under consideration is delineated, a boundary smoothing process is applied and the row detection process begins. Row detection operates by means of an active contour model based on a network of parallel lines. The last step is the design of vegetative vigor maps. Row vigor is computed using pixels neighboring the lines of the network. Once row vigor is obtained on the rows, 2D vigor-maps are constructed. The values measured on the row are propagated to the inter-row pixels, producing «continuous» vigor maps ready to be exported to a GIS software. We successfully exercised our framework on vineyard images. The resulting parcel segmentations and row detections were accurate and the overall computational time was acceptable.

Download the PDF

DOI:

Publication date: December 22, 2021

Issue: Terroir 2006

Type: Article

Authors

Jean-Pierre DA COSTA, Christian GERMAIN, Olivier LAVIALLE, Saeid HOMAYOUNI and Gilbert GRENIER

LAPS CNRS – ENITAB – ENSEIRB, Université Bordeaux 1
351 cours de La Libération, 3305 Talence cedex, France

Contact the author

Keywords

remote sensing, image processing, row crop, vine

Tags

IVES Conference Series | Terroir 2006

Citation

Related articles…

Étude intégrée et allégée des terroirs viticoles en Anjou: caractérisation et zonage de l’unité terroir de base, en relation avec une enquête parcellaire

The terroir concept is presented as the basis of the A.O.C system, in the french vineyards. The “Anjou terroirs” programme aims at bringing the necessary scientific basisfor a rational and reasoned exploitation of the terroir. lt must lead to finalizing a lighter, more relevant integrated method of characterisation wich could be generally applied.
Read More

Climate change impacts on European grapevine yields through a dynamic crop modelling approach

Climate has a predominant role on growth and development of grapevines. Therefore, climate change represents an important challenge to the winemaking sector.

Read More

Chinese localization of wine aroma descriptors

Wine aroma descriptors are important tools for wine evaluation. The present well-known wine aroma descriptor system was created and based on Western culture, which makes difficult for Chinese consumers to recognize and learn wine. AIM: The aim of this study was to update the wine aroma descriptor system for Chinese.

Methods: Fifty-four wine aroma descriptors of ‘Le nez du vin’ was used as substitution candidates. Firstly, a survey on unfamiliar aromas was distributed to 150 untrained Chinese wine consumers. Twenty attributors, such as blackcurrent buds, quince, linden, were selected as the most 17 unfamiliar. Then, a descriptive analysis was performed by trained tasting panel to substitute the targeted twenty aromas perfume. Furthermore, reference standards were looked and new le nez du vin were made. Finally, a substitution analysis was performed to replace the unknown wine aroma to the Chinese local aromas.

Read More

The future of pesticide regulation in the EU – between precaution and proportionality

The article analyzes current developments in European pesticide regulation.

Read More

Prediction of aromatic attributes of red wines from its colour properties 

Wine perception is a multisensory experience that makes use of the sight, smell, and taste senses. When wine is sensorially assessed, the stimulus received generates multiple signals that tasters convert into organoleptic descriptors. Colour is commonly the first attribute evaluated during wine tasting. Moreover, the colour properties provide the taster with a priori information of the wine’s aroma. This preconceived perception is later confirmed or denied during the aroma evaluation.

Read More