A procedure for monitoring the phenological status of peach flowers with artificial vision

Hansen, Mark; Veganzones, Alvero; Lafuente, Virtoria; Barreiro, Pilar; Lleo, Lourdes; Val, Jesus

Abstract

Tree flowering is a major event in crop production as it anticipates season yield. However a number of issues may occur during the campaign such as frost, and/or irregular mineral nutrition, among other, that strongly affect this process. On the other hand many fruit species show the phenomenon of “vecería” that refers to the fact that the trees have an increased yield every two year. Therefore, fruit growers and production engineers demand an insitu tool that would allow providing quantitative features regarding the amount of flowering, their phenological status, or even the presence of strong mineral deficiencies which lead to an abnormal development of the sexual organs of the flowers. Artificial vision suites into such a demands since it can be performed in the fields on-board of drones. In this work free online tools such as Makesense, Roboflow and YOLOv4 have been used to train and validate an automated procedure for the identification of the position of the sexual organs of Caterina Peach Flowers in order to relate their status with the nutrition state of the trees.

• Presentation Conference Type: Conference Paper (unpublished)
• Conference Name: The XX CIGR World Congress 2022
• Start Date: Dec 5, 2022
• End Date: Dec 10, 2022
• Deposit Date: Dec 13, 2022
• Keywords: Phenology, Phenological status, Peach flowers, Artificial vision
• Public URL: https://uwe-repository.worktribe.com/output/10230255
• Related Public URLs: http://cigr2022.org/