Tomato leaf diseases classification using alexnet

Abstract

The economy is largely reliant on agricultural productivity. Plant diseases and pests are also a serious issue in agriculture. It is necessary to recognise them early on in order to eliminate all infections as rapidly as possible and avoid crop devastation. To protect the plants against illness, a variety of insecticides have been utilised. Despite these precautions, the illness continues to spread throughout the field. Because we don't always know what kind of sickness we're dealing with, a bad pesticide may have been applied instead. As a result, it's all futile.Tomato leaf infections, on the other hand, are a severe problem for many farmers, thus mastering the severity of diseases in a fast and precise manner is critical to assisting staff in taking additional intervention steps to prevent plants from becoming more afflicted. An Alexnet model for detecting tomato leaf disease is proposed in this study, which may minimise the number of training processes while enhancing computation accuracy and gradient flow. This indicates that categorization of diseases is equally important. An improved classification model is presented in this paper for identifying and categorising tomato leaf disease. Before CNN is used to identify the pictures, a training dataset containing a large number of images is utilised, and visual characteristics are extracted using several methods. The Alexnet model achieves the highest classification accuracy when compared to other models.