Plant Disease Identification and Detection Using Machine Learning Algorithms

Abstract

Plant diseases cause significant reductions in plant production, leading to economic losses. Pests and diseases destroy plants or parts of plants, leading to lower agricultural production and food insecurity. Pathogen and pest infestations are major causes of crop destruction and loss of crop yields as most farmers do not have access to pathologists and entomologists or sufficient information and infrastructure to identify the disease early on and reduce the loss. To address this, we used a plant leaf dataset containing healthy and diseased plant leave images. These data sets are freely and publicly available to be used. And by using Machine learning and Computer Vision to identify to train the classifier, classify and predict the health of the plant. Using this method, we detect different kinds of disease present in the plant at an early stage before it spreads on a higher scale and a large part of the field is infected and accordingly take action to reduce the amount of loss. The Random Forest model used in this study achieves an overall accuracy rate of 95% on the specified test images. With a day-by-day increase in handheld devices, this can be a successful approach in helping the food industry to cope with the need for food for the rapidly growing population of the world.

Topics & Concepts