Ascribing Machine Learning Classifiers to diagnose the attacks of Alternaria solani on Leaves of Solanum tuberosum

Abstract

Following the recent advances in technology, came advanced computational domains like, Internet of Things, Machine Learning, Artificial Intelligence, Data Science, and many more. These fields really tend to help mankind a lot. In this work, we would make use of Machine Learning aspects to perform prediction of diseases in plant. Specifically, spot the Early Blight Disease in the Potato Leaves. The potato plant, Solanum tuberosum, is a significant crop that is grown all over the world and generates large quantities of tubers that are a good source of nutrients. The potato has many medicinal benefits in addition to being a common staple diet. When the fluid out from tubers is consumed in moderation, it can treat gastric ulcers and relieve inflammation and acidity. Two harmful potato diseases, late blight and early blight, are pervasive. Everywhere potatoes are cultivated, both are present. The labels “Early” and “Late” allude to the relative timing of their field emergence, however both disorders might manifest simultaneously. In this work, we would focus on Early Blight. The fungus Alternaria solani, that can infect potatoes, tomatoes, several species of the potato genus, and some mustards, is the cause of early blight of potatoes. Young, actively growing plants are rarely impacted by this disease, commonly known as target spot. It first appears on elder leaves. Warm temperatures and heavy humidity foster Early Blight. This disease affects the tuber symptomized by dark, rounded to irregular dots being developed on the tuber. As the disease develops, the flesh of the potatoes commonly becomes water-soaked yellow to greenish yellow. For this work, we have collected a set of nearly 1000 samples of Early Blight affected Potato Leaves. Using that, we have modelled a Machine Learning Classification Paradigm that could potentially predict the occurrence of Early Blight Disease making using of classical classifier algorithms. Medical Science have advanced to a great height. If we could potentially predict the disease, in the early stages, Plant Pathology could stop the menace from occurrence.