Moving beyond visible flower counting: RGB image-based flower number and yield prediction in grapevine

Accurate yield estimation is crucial for optimising vineyard management and logistical organisation. Traditional methods relying on manual and destructive flower or berry counts are labour-intensive and unsuitable for large-scale applications. To address these limitations, this study evaluated the potential of RGB image analysis for non-destructive estimation of flower number in grapevine inflorescences and its relationship with yield components. The objectives were: (i) to correlate the number of actual flowers with the number of pixels in inflorescence images, considering the projected area and not visible flower count; (ii) to assess the model’s ability for predicting yield components (number of berries and bunch weight) on the plant scale. The study was conducted during the 2023 vintage in a cv. Catarratto/1103P vineyard located in Sicily, on 36 vines. The vines were trained on a vertical shoot positioning trellis system with a bilateral spur pruned cordon. All the inflorescence images were acquired using a smartphone and white cardboard under field conditions. Images were analysed using FIJI/ImageJ© software, inflorescences were segmented via Otsu’s method, and pixel counts were used to estimate flower number through linear regression. Flower counts were validated through manual counts of detached calyptra. Vines were harvested to measure bunch weight, and the number of berries was counted to calculate the fruit set rate. External model validation was performed on datasets from Catarratto, Chardonnay, and Vermentino cvs. Results showed a strong correlation between inflorescence pixel count and flower number, both on single inflorescences (R2 = 0.78; MAPE 29 %; n = 300) and in terms of total flowers per vine (R2 = 0.95; MAPE 12 %), with estimation accuracy slightly varying according to inflorescence length. External validation yielded an MAPE of 20 % in Catarratto, 40 % in Chardonnay, and 34 % in Vermentino. The model also reliably predicted the number of berries (R2 = 0.94; MAPE 13 %) and bunch weight (R2 = 0.79; MAPE 20 %) per vine. While environmental factors such as fruit set can affect yield, this study highlights the potential for early yield prediction with image analysis. The methodology holds promise for scalable applications and integration into vineyard management technologies in wine and table grapes.