Polynomial relation between mean shoot area and mean shoot weight for vigour calculation on cv. Touriga Nacional

Abstract

Aim: To develop a new methodology for determining vine vigour based on a polynomial relation between mean shoot area and mean shoot weight. Methods and results: Thirty grapevines (cv. Touriga Nacional) characterized by low, medium and high vigour were selected for a three-year study. Vine vigour was determined each year by using the traditional “mean shoot weight” methodology. At the same time, the mean shoot area of each vine was determined by using digital image processing, where RGB images (Red Green Blue) were converted to HSV images (Hue Saturation Value) and then to binary images based on the saturation values with a threshold set to 0.15. Finally, we established a polynomial relation between mean shoot area and mean shoot weight, given by the equation f(x) = 0.0047x2 + 0.2254x + 0.0474, where f(x) corresponds to the mean shoot weight (vine vigour estimation) and x corresponds to the mean shoot area. Conclusions: The proposed method for calculating vine vigour based on a digital image of a vine and a polynomial relation between mean shoot area and mean shoot weight provides a highly reliable estimate of the actual vigour. Correlation for the 30 grapevines with three-year average of 99.8%. Significance and impact of the study: With this new methodology, it is unnecessary to determine the pruning weight and the number of shoots per plant in the vineyard. Its implementation is simple, based on a digital photography of the vine under study, and independent of the plot pruning date.

Keywords

• vine
• vigour
• digital image processing