Plant Disease Detection Using Machine Learning

Crop diseases are a significant risk to our capacity to grow sufficient food. Although speedy identification is essential, it usually becomes difficult due to the unavailability of resources in most of the world. Luckily, recent developments in analyzing leaf images have been quite successful. This paper explores the ability of the Random Forest machine learning algorithm to distinguish between healthy and infected plant leaves based on image data. Our approach is a step-by-step process: first, generating a targeted dataset; second, deriving prominent visual features from the images; third, training the Random Forest classifier; and lastly, executing the classification. By having the model learn from a dataset of images of healthy and sick leaves, and using the Histogram of Oriented Gradients (HOG) to extract image features, this research shows how machine learning can be used to facilitate mass plant disease detection with publicly available information.