Crop Disease Detection Using YOLO

摘要

Agriculture is the cumulative activity for millions of farmers in India. Planters have a wide range of diversity for selecting suitable crops. But due to scarcity of knowledge, farmers are in a daze about kinds of diseases that affect the farm. Many farmers struggle and waste much of their time in reaping diseased crops. The timely assessment of the problem is necessary to avert major damage and enhance production. The proposed system makes use of a novel approach of the object detection technique to detect plant disease, YOLO(You Only Look Once). YOLO processes leaf images at 45 frames per second in real-time, which is faster than other object detection techniques. It divides the image into several grid cells before processing the image. The bounding boxes and class probabilities are predicted by a single neural network in just one evaluation. This effectively boosts the speed and accuracy of disease detection on the leaf.