Leaf disease detection is a critical task in precision agriculture, aiming to monitor and control the spread of plant diseases for sustainable crop management. Object detection models have shown promise in accurately identifying and localizing diseases on plant leaves in recent years. This paper explores the effectiveness of YOLOv3 (You Only Look Once) and a variant known as Gaussian YOLOv3 in the context of leaf disease detection. YOLOv3 is known for its real-time object detection capabilities and high accuracy. However, it may face challenges in accurately localizing subtle disease patterns and handling uncertainties in complex leaf images. To address these challenges, Gaussian YOLOv3 incorporates Gaussian components to model uncertainty and improves localization accuracy. The comparative analysis involves evaluating the performance of YOLOv3 and Gaussian YOLOv3 in terms of localization accuracy, speed, adaptability to diverse conditions, and training requirements. Experiments are conducted using a dataset comprising various leaf diseases under different environmental conditions. They enable timely interventions and agricultural decision-making, reducing crop losses and ensuring effective disease management.