Commercial Plant Leaf Disease Detection in the Coastal and Malnad Regions of Karnataka using CNN, DenseNet121, and InceptionV3

Leaf diseases pose a major threat to the productivity and quality of commercial crops in the coastal and Malnad regions, renowned for their diverse and high-value agricultural practices. In this work, a sophisticated deep learning modelbased automated method for identifying and categorising leaf diseases is presented. We analysed photos of both healthy and diseased leaves from important crops in these areas using three models: Convolutional Neural Networks (CNN), InceptionV3, and DenseNet121. The CNN model was custom-built to address the specific challenges of disease detection, while InceptionV3 and DenseNet121, pre-trained on large datasets, were fine-tuned for improved performance on the domain-specific dataset. The models demonstrated high accuracy in identifying diseases, with DenseNet121 excelling in detecting intricate patterns and InceptionV3 ensuring efficient feature extraction. The proposed system enables rapid and accurate diagnosis, aiding farmers in timely interventions to mitigate crop losses. The findings highlight how deep learning may be used to solve problems in agriculture and provide a scalable and affordable way to enhance crop health. By supporting sustainable farming methods, this project improves resilience and productivity in the Malnad and coastal areas.