Fruit Quality Detection:A Comparative Study of the YOLO Series

Fruit Quality detection is an important aspect in the agricultural and food supply chain, it is necessary to ensure that only fruits that meet a certain standard of quality reach consumers. Generally, this process is done manually by workers carrying out visual inspection based on the fruit's appearance, texture, and color; this can be time-consuming and is subject to human errors. This may cause inefficiencies in large-scale production and possibly compromise the quality of fruits reaching the market. A fruit quality detection system is designed in this study using computer vision techniques, and a comparative analysis of YOLO series models is conducted for fruit quality detection. The analysis focuses on various performance metrics, including speed, accuracy, and robustness, while considering the benefits of each architecture. Each model is assessed to highlight its key strengths and ability to accurately classify fruit quality, with the goal of selecting the most suitable model to address the shortcomings of traditional methods. The performances of YOLOv8, YOLOv9, YOLOv10, and YOLOv11 are analyzed in terms of precision, recall, and mean Average Precision (mAP). YOLOv9 had the highest mAP50 and mAP50-95 with the score of 0.981 for both. Furthermore, the model has a very high recall of about 0.936, which makes it ideal for detection tasks needed in precision-demanding areas.