Classification of Corn Leaf Disease using Resnet18,Alexnet and VGG16

Corn leaf disease detection models aim to accurately identify and diagnose diseases in crops, enabling early intervention to prevent spread and minimize losses. By providing timely and precise information, these models help farmers reduce the use of chemical inputs, increase productivity, and promote sustainable agricultural practices. The ultimate goal is to enhance crop health, yield, and profitability while minimizing environmental impact. To achieve the specified goals in corn leaves, a Corn Leafe Disease Detection Model has been developed utilizing three distinct algorithms: ResNet18, VGG16, and AlexNet. Algorithms were carefully chosen due to their effective performance in similar tasks. The model was constructed using a pre-trained dataset obtained from Kaggle website consisting of images of corn leaves with disease as well as healthy leaves, ensuring a robust foundation for accurate disease detection in corn crops. The Kaggle website provided a pre-trained dataset with pictures of both healthy and infected corn leaves, which was used to build the model. This dataset ensured a robust foundation for accurate disease detection in corn crops. The dataset underwent training, validation, and testing using deep learning algorithms. Results were analyzed based on accuracy and loss measures to assess the performance of the model. A number of scholars have used deep learning and machine learning approaches to create models for identifying crop conditions. However, the comparison conducted in this paper primarily focuses on the results of validation testing to make decisions regarding the selection of algorithms for detection purposes. Among the three algorithms evaluated, VGG16 has consistently demonstrated high performance determined by precision, loss, confusion matrix, precision, recall and F1 score.