Abstract—Plant diseases must be identified early to protect crop harvests, as agriculture plays a crucial role in ensuring global food security. This paper introduces an advanced deep-learning approach utilizing a conventional Convolutional Neural Network (CNN) for the multiclass classification of groundnut leaf diseases. The research focuses on constructing a robust deep learning model, named Groundnut Leaf Disease Identification Classification Convolution Neural Network (GLDICCNN), to rapidly identify, classify, and predict groundnut leaf diseases. The methodology encompasses comprehensive data collection from agricultural fields, preprocessing, model development, and rigorous evaluation. The proposed model, Groundnut Leaf Diseases Identification, Classification with Convolution Neural Network (GLDICCNN) demonstrates impressive performance metrics after extensive experimentation. The training accuracy reaches 99.73%, while validation accuracy stands at 97.06%. Correspondingly, the training and validation loss values are 0.0035 and 0.1649, respectively. Evaluation metrics, including precision (96%), recall (96%), and F1Score (96%), highlight the effectiveness of the proposed model. Moreover, the test accuracy attains a commendable 96%. Comparative analysis with pre-trained models such as ResNet50, ResNet101, and ResNet152 underscores the superior accuracy achieved by the GLDICCNN model. In summary, this research establishes a deep learning framework that excels in groundnut leaf disease identification, classification, and prediction. The findings underscore the potential of the proposed model for practical applications in agriculture, contributing to enhanced crop yield protection and global food security.
 
Keywords—agriculture, augmentation, deep learning, groundnut leaf, hyperparameter tuning, Residual Neural Network (ResNet) models, transfer learning


Copyright © 2024 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.