Plant Disease Detection And Treatment Using AI For Modern Agriculture

Abstract

Plants diseases are one of the hardest challenges in world agriculture where it has traditionally been diagnosed by farmers or experts that perform manual inspection. This traditional method is slow, subjective and erratic, resulting in delays in treatment and considerable loss of crops. With the advent of artificial intelligence, new applications nowadays allow automated, data-driven solutions to optimize agricultural accuracy and efficiency. Here we report an ongoing problem of disease identification being too late and not always accurate due to relying on visual observation and having few experts for consultation. A solution to these challenges is considered in the proposed system, which utilizes an AI/MLbased model to identify plant disease right from leaf images. The procedure involves application of image pre-processing, feature extraction and deep learning classification algorithm using Convolutional Neural Networks(CNNs) to successfully detect disease types with higher accuracy. Moreover, the system is also equipped with a treatment-recommendation block that associates each determined disease to appropriate cures, providing farmers with an immediate actionable advice. The results indicate that the model can efficiently distinguish healthy leaves and ill ones across various plant species with good accuracy rate; enabling to give a prompt diagnosis and treatment if necessary. With this combination significant reduction of human dependency is achieved and the burden on the crop damage is also reduced, leading to sustainable technology-based farming.