A Comprehensive Analysis of Plant Disease Detection Using Advanced CNN Models

Abstract

Plant diseases are one of the primary difficulties encountered by farmers and farmers in the globe. Plant disease identification is vital in handling the care of plants. This paper describes a method for identifying plant diseases using images of their leaves that is based on Convolutional Neural Networks (CNNs). There are a total of four categories here, they are healthy, rust, powdery, and blight plants. Approximately 2123 photos were utilized for training testing and validation purposes. This research evaluates the usage of sophisticated convolutional neural network models, especially, VGG19 and ResNet, in the identification of plant diseases. The study underlines the superiority of VGG-19, indicating its potential for accurate and reliable plant disease diagnosis, while also revealing insights into areas for improvement in plant disease identification using image data. VGG19 developed a model with 99.35% acuuracy which is deemed higher than the other two models findings will be important for establishing dependable and precise plant disease detection systems and setting the bar for precision farming and sustainable agricultural production.