REAL-TIME MECHANICAL FLOWER THINNING EQUIPMENT, CONTROLLED BY ARTIFICIAL INTELLIGENCE

In this paper, the designing and development of a novel mechanical flower thinning equipment, destined to increase the fruit production in orchards, is presented. The system integrates a ZED 3D camera with a dedicated controller for artificial intelligence running a custom trained YOLO9 algorithm, for real-time flower detection and counting. Based on the flower density data, the rotational speed of the thinning rotor is automatically adjusted to achieve the desired thinning ratio. Laboratory tests were conducted to evaluate the efficiency and adaptability of the YOLO9 algorithm to control the equipment in simulated flower density conditions. Results demonstrated potential improvements in thinning accuracy, contributing to optimized fruit development, and reduced manual labor. The proposed equipment offers an innovative approach to orchard management, ensuring sustainable practices by enhancing flower thinning precision while reducing labor costs.