Using deep learning for pruning region detection and plant organ segmentation in dormant spur-pruned grapevines.

Even though mechanization has dramatically decreased labor requirements, vineyard management costs are still affected by selective operations such as winter pruning. Robotic solutions are becoming more common in agriculture, however, few studies have focused on grapevines. This work aims at fine-tuning and testing two different deep neural networks for: (i) detecting pruning regions (PRs), and (ii) performing organ segmentation of spur-pruned dormant grapevines. The Faster R-CNN network was fine-tuned using 1215 RGB images collected in different vineyards and annotated through bounding boxes. The network was tested on 232 RGB images, PRs were categorized by wood type (W), orientation (Or) and visibility (V), and performance metrics were calculated. PR detection was dramatically affected by visibility. Highest detection was associated with visible intermediate complex spurs in Merlot (0.97), while most represented coplanar simple spurs allowed a 74% detection rate. The Mask R-CNN network was trained for grapevine organs (GOs) segmentation by using 119 RGB images annotated by distinguishing 5 classes (cordon, arm, spur, cane and node). The network was tested on 60 RGB images of light pruned (LP), shoot-thinned (ST) and unthinned control (C) grapevines. Nodes were the best segmented GOs (0.88) and general recall was higher for ST (0.85) compared to C (0.80) confirming the role of canopy management in improving performances of hi-tech solutions based on artificial intelligence. The two fine-tuned and tested networks are part of a larger control framework that is under development for autonomous winter pruning of grapevines. Supplementary Information: The online version contains supplementary material available at 10.1007/s11119-023-10006-y.

Canopy physiology, vine performance and host-pathogen interaction in a fungi resistant cv. Sangiovese x Bianca accession vs. a susceptible clone.

The present study compares the physiological and cropping response of the new fungi-resistant grapevine Accession 72-096 ('Sangiovese' x 'Bianca' hybrid) against a susceptible 'Sangiovese' clone which was either fully (FS-SG) or partially sprayed (PS-SG). Data logged on Accession 72-096 indicate that while two early season sprays were enough to avoid major downy mildew (DM) and powdery mildew (PM) outbreaks, Accession 72-096 also showed concurrent desirable features such as moderate cropping, loose clusters, fast sugar accumulation coupled with sufficient acidity even at peak total soluble solids (TSS) concentration (around 24 °Brix), good color and higher flavonols prompting co-pigmentation. Conversely, FS-SG showed final lower acidity despite the notably lower sugar concentration (≅18 °Brix), as well as larger clusters and berries that resulted in more compact bunches. From a methodological viewpoint, end of season single-leaf readings appeared to overestimate the limitation of leaf function due to PM and DM infections in SG-PS vines which, when assessed via a whole-canopy approach, did not show significant differences vs. Accession 72-096, a result likely due to counteracting effects linked to a compensation mechanism by healthy tissues. Our data also suggest that a PM infection can lead to a decoupling in sugar-color accumulation patterns.