Application of the YOLOv8 Model to a Fruit Picking Robot

Due to the current labor shortage in fruit picking, the utilization of robots for harvesting has become crucial in addressing this issue. This paper presents two key models, an apple image recognition model and an apple image recognition classification model, both based on Yolov8. These models are specifically designed and trained to be implemented in apple picking robots. In Model1, the RGB color space was chosen, and the image data underwent preprocessing using Gaussian filtering. The training process utilized manually labeled data. Model 2, on the other hand, involved data grouping and training using the Yolov8 model, with a focus on optimizing the recognition results. The feasibility of these models was rigorously tested through extensive modeling experiments. This research not only offers a solution for the labor shortage in fruit picking robots but also demonstrates its wide-ranging potential in real-life applications and sales processes. Furthermore, the insights gained from this study can contribute to the advancement of automatic recognition and classification of other fruits, thereby fostering the intelligent development of fruit picking robots.