Implementation of Fertilizer Recommendation System for Disease Prediction Using Deep Learning Algorithm

Abstract

In addition to providing food for a growing population, agriculture also helps to combat global warming and is a source of energy. Because they may have a detrimental effect on the type and quantity of produced in agriculture, plant diseases are particularly significant. Identification of plant diseases at an early stage is essential for treatment and crops disease management. To identify ailments, people frequently use their naked eyes. This technique involves experts with the capacity to spot variations in leaf colour. This process is labor-intensive, time-consuming, and unsuitable for large fields. Multiple medical experts would frequently give various diagnoses for the same illness. Due to the ongoing need for professional supervision, this technology is expensive. Plant diseases can increase the cost of agricultural output and, if left untreated, can completely bankrupt a company. To control the spread of a plant disease at a low cost and save the bulk of the harvest, growers must keep an eye on their crops and be able to identify early indications. Expert agriculturists may be excessively expensive to hire, especially in rural and inaccessible regions. An alternative way to plant monitoring can be provided by a deep learning algorithm embedded in a picture, and this method can be managed by a specialist to offer less expensive services. It includes features extraction and classification, as well as an image classification method that anticipates several illnesses in maize leaf using a neural network approach. Additionally, adapt the method to prescribe fertilizers based on severity analysis and data.