Deep Learning-Based Cotton Plant Disease Detection Using CNNs: A Smart Agriculture Approach

Abstract

The increasing prevalence of plant diseases poses a significant threat to global cotton production, leading to substantial economic losses and reduced crop yield. Traditional manual disease detection methods are time-consuming, labour-intensive, and often inaccurate. To address this challenge, this research proposes an advanced deep learning-based approach for automated cotton plant disease detection using Convolutional Neural Networks (CNNs). The study evaluates multiple CNN architectures, including GoogleNet, VGG16, DenseNet201, ResNet50, and TLResnet152V2, to determine their effectiveness in identifying and classifying diseased cotton leaves. The proposed methodology leverages normalized and augmented datasets, utilizing data pre-processing, feature extraction, and transfer learning techniques to enhance model performance. Extensive experimental evaluations demonstrate that data augmentation significantly improves classification accuracy, enabling CNN models to generalize better across diverse disease conditions. Among the tested architectures, TLResnet152V2 achieved the highest accuracy (92.03%) and F1-score (0.8842), outperforming all other models, followed closely by ResNet50. These results highlight the superiority of deep residual learning in plant disease classification, ensuring robust feature extraction and precise detection. This studies also explores the combination of CNN-primarily based disorder detection into clever agriculture structures, allowing actual-time sickness classification via cell packages and IoT-based totally answers. The findings affirm that deep gaining knowledge of-pushed plant disorder detection can considerably enhance precision farming, reducing dependency on professional agronomists while improving early disorder intervention techniques. destiny studies will awareness on deploying light-weight CNN models for facet computing, integrating climate statistics for predictive disorder modelling, and exploring hybrid deep studying strategies for enhanced accuracy. The examine demonstrates that CNN-based automatic cotton plant disease detection is a transformative step closer to sustainable, AI-enabled smart agriculture, ensuring better productivity, decreased crop losses, and advanced food safety.