Machine Learning Models for Plant Disease Prediction and Detection: A Review

Abstract

In the agriculture field farmers are dependent on crops and once it is caused by disease, they can lose the proper yield. Plant illnesses are the common cause of low yields and reduced income for farmers. So, to overcome this, machine learning approach in the agriculture field is showing exponential improvement for interdisciplinary research, but in modern disease prediction systems, it is a different process that can takes more time to identify and understand the type of disease. Disease prediction is related to computer vision and machine learning to detect types of leaf disease on different plants. It’s a long-term system that includes a variety of real-world scenarios such as, Prediction system for vine leaf disease, pomegranate leaf disease, corn leaf disease, etc. Disease prediction is the accurate determination of the disease state of plant leaves. Accurate disease prediction is one of the key requirements in data science. Presently, machine learning based methods have improved prediction accuracy for plant leafs like grapes, pomegranates, maize. However, the disease prediction performance is still required to be improved in this challenging environment. Existing disease prediction models require high computational time and storage facilities in the agriculture field. To overcome this, we have proposed a comparative study of ML for prediction and classification of crop diseases to improve the efficiency of early prediction of crop diseases.