Vineyard yield estimation using image analysis: assessing bunch occlusions and its dependency on fruiting zone canopy features
Abstract
Performing accurate vineyard yield estimation is of upmost importance as it provides important benefits to the whole vine and wine industry. Recently, image-analysis approaches have been explored to address this issue however this approach has as main challenge the bunch occlusion, mostly by vegetation but also by neighboring bunches. The present work aims at assessing the magnitude of bunch occlusion by neighboring bunches and to evaluate its dependency on a selection of vegetative and reproductive vine parameters assessed at fruiting zone. Forty vine segments (1 m) of two vineyard plots of the white cultivars ‘Alvarinho’ and ‘Arinto’ were assessed for vegetative and reproductive features at fruiting zone and imaged with a 2D camera. Bunch occlusion by leaves presented the highest occlusion rates, reaching an average of 68.5% for both varieties. Bunch occlusion by neighboring bunches presented average values of 12.2 and 15.2%, respectively for ‘Alvarinho’ and ‘Arinto’. Regarding the correlations between the rate of bunch-by-bunch occlusion and the assessed vegetative and reproductive variables, all correlation coefficients were non-significant, indicating that this type of occlusion is not driven by one single variable but, instead, by an interplay of factors that integrates the effects of several vegetative and reproductive canopy features. The magnitude of bunch-by-bunch occlusion also shows that this type of occlusion shouldn´t be neglected as it can induce an underestimation of the yield, mainly when the image-analysis algorithms are based on the relationships between visible bunch area and bunch mass.
DOI:
Issue: ICGWS 2023
Type: Poster
Authors
LEAF—Linking Landscape, Environment, Agriculture and Food Research Center, Associated Laboratory TERRA, Instituto Superior de Agronomia, Universidade de Lisboa, Tapada da Ajuda, 1349-017 Lisboa, Portugal
Contact the author*
Keywords
Grapevine yield prediction; proximal sensing; bunch-by-bunch occlusion; Vitis vinifera L.