Developed a Tomato-Condition Classification System using Image Processing and YOLO Technology

Classifying tomatoes after harvesting still heavily relies on manual labor, leading to inefficiencies and inconsistent quality. This issue affects the cost and reduces the ability to export to important international markets. This study introduces a modern solution by applying the YOLO algorithm to automate the tomato classification process. The automatic classification system is proposed, which not only reduces cost labor but also improves performance reaching approximately $98 \%$ on TOMATO dataset. This system utilizes image processing technology to assess size, color (including ripe, unripe, and rotten tomatoes), and shape (recognizing abnormalities, surface scratches,…). Additionally, the research emphasizes the importance of applying computer vision technology and utilizing self-created datasets to ensure high performance in the real-world environment of tomatoes, thus opening up prospects for technology application in agriculture.