| dc.description.abstract | Precision pomology is the application of site-specific management into orchards. Orchards produce high value crops and the application of site-specific management has high potential. Finding plant or soil characteristics that would correlate or predict yield variability would be of great importance for applying site specific management. Such data often come from remote sensing plant indices. Flowering distribution in the field has been suggested as a possible early in the season characteristic that can lead to successful site specific management for the rest of the year. Our objectives were (1) to study the flower density variability in an apple orchard and (2) to find a quantitative variable related to an image depicting a tree in full blooming, that is correlated with the flower density and consequently to the fruit yield of the tree in the current year. The research was carried out in a commercial 5 ha apple orchard, located in Agia area, Central Greece. The orchard included two apple cultivars, Red Chief (main cultivar) and Golden Delicious (pollinator). The between-row spacing of the trees was 3.5 m and the intra-row 2 m. Trees were trained as free palmette. In April 2007, when the trees were in full bloom, 200 photos of whole trees of Red Chief cultivar were taken in a grid of 20×7 m. The location of those trees was recorded using a hand-held computer with GPS in order to create the flower map. In September 2007 yield mapping was carried out measuring yield per ten trees and recording the position in the centre of the ten trees. Using this data, e.g. the measured yield of the trees and their corresponding images, an image processing based algorithm was developed that predicts the yield of a tree by analyzing its image when it is in full bloom. For the evaluation of the algorithm a case study scenario is presented where the error of the predicted yield was found to be in the order of 18%. These results indicated that potential yield might be predicted early in the season from flowering maps, which is very important for the farmer and fruit industry. Other potential uses of the flower map include planning site-specific chemical thinning, fertilization and application of other cultural practices in the orchard based on flower density variability. | en |
