Development of Tomato Leaf Disease Detection using YoloV8 Model via RoboFlow 2.0

The Philippines, as a tropical archipelago with 30% of its land classified as agricultural, produces tomatoes throughout the year, though peaks during dry season. Not accounting its climate and weather conditions and geographical challenges, pest or diseases infestation is one of the hindrances in maximizing the production, yield, and shell life of tomato harvest. This can be mitigated by early prognosis though its leaf. The use of YoloV8 via RoboFlow is presented in this paper to detect nine (9) common tomato leaf diseases. It generated 99% average precision from both the validation and test sets. YoloV8 model has proven its high potential to classify tomato leaf disease when it resulted in a mean Average Precision (mAP) of 98.9%, at an average precision rate of 97.5% and recall rate of 91.9%.