By Franka Gabler
The 2018 Efficient Vineyard table grape work began in April, when the Carnegie Mellon University research team visited the Flame Seedless vineyard in Delano, Califonia. The first trip coincided with inflorescence emergence – the goal was to image, detect, and count developing clusters, since inflorescences are less occluded by developing foliage at this stage. Initial inflorescence count can serve as an early predictor of a vineyard crop size. The other goal was to image the developing canopy and extract from the images the information on the number of shoots, and/or exposed leaf area, since a canopy is conducive for this work at this early stage of development. The second trip was made in late April, at a later stage of canopy development, with a goal to image and estimate exposed leaf area, both in Flame Seedless and Crimson Seedless vineyards.
Shoot count and exposed leaf area may be suitable methods to estimate vine vigor, and may correlate with canopy size during later grapevine annual development. As inner canopy layers develop, it becomes difficult to accurately estimate vine vigor, since inner canopy layers become hidden. In addition, following fruit set, growers start managing the canopy by removing leaves and cutting shoots growing into an adjacent row trellis, which can mask true vine vigor – if canopy size is estimated at this time.
The researchers will investigate the relationship between early canopy size and visible fruit at veraison/harvest. If a relationship exists, this could be used to improve the algorithm for yield estimation by incorporating a variable fruit occlusion factor, based on canopy size.