Pineapple Ripeness Detection Using YOLO v8 Algorithm

Abstract

Pineapple is a tropical fruit that has various benefits for human health. However, determining the ripeness of pineapple is not easy, especially for inexperienced consumers. This study aims to develop a computer vision system that can automatically identify the ripeness of pineapple based on its peel color and texture. We use the YOLO v8 algorithm, a state-of-the-art object detection model, to detect and classify pineapple images into three categories: unripe, ripe, and overripe. We collect and label our own dataset of 1386 pineapple images with different ripeness levels. We train and test the YOLO v8 model on our dataset using Google Colab. We evaluate the model’s performance using accuracy. The results show that the YOLO v8 model achieves an accuracy of 84.75%. These results indicate that the YOLO v8 model can effectively detect and classify pineapple ripeness with high accuracy and speed. This system can be useful for consumers, farmers, and retailers to select and purchase high-quality pineapples.