Innovative Approaches to Tomato Leaf Disease Detection Bridging Tradition and Technology.

Abstract

This study explores the transformative potential of integrating Artificial Intelligence (AI) with precision agriculture to address key challenges in farming, such as disease detection, resource optimization, and yield improvement. By leveraging the Real-Time Detection Transformer (RT-DETR) framework, the study combines high-resolution imaging, hyperspectral scanners, and real-time data processing to enable efficient and accurate detection of tomato leaf diseases. A robotic system, equipped with autonomous navigation and non-invasive diagnostic tools, was developed to classify diseases and provide actionable insights for sustainable crop management. The study evaluates the performance of object detection models, such as YOLOv8 and RT-DETR, on the PlantVillage and PlantDoc datasets. RT-DETR achieving a mean Average Precision (mAP) of 0.988 on the controlled PlantVillage dataset and 0.402 on the complex PlantDoc dataset. By integrating AI with farming practices, this research highlights a pathway to improving agricultural productivity, environmental resilience, and global food security.