Revolutionizing Arecanut Disease Detection through Deep Learning Techniques

Arecanut, also known as betel nut, is a significant tropical crop, with India ranking second globally in both production and consumption. However, its lifecycle is threatened by various diseases, impacting everything from roots to fruits. Currently, disease detection relies on manual observation, requiring farmers to painstakingly inspect each crop at intervals. In this study, we introduce a novel system that leverages Deep Learning techniques, specifically Residual Networks (ResNet), to identify diseases in arecanut leaves and trunks, and provide corresponding remedies. ResNet, an advanced variant of Convolutional Neural Networks (CNNs), excels at learning intricate features through residual blocks, enabling the creation of deeper model architectures without encountering vanishing gradients. We curated a bespoke dataset comprising 620 images encompassing healthy and diseased arecanuts for training and testing the ResNet model. The data split followed an 80:20 ratio for training and testing, respectively. For model compilation, we employed categorical cross-entropy as the loss function, adam as the optimizer, and accuracy as the evaluation metric. The training process spanned 50 epochs to achieve optimal validation and test accuracies while minimizing loss. Our approach yielded promising results, showcasing an accuracy of 94.46\% in accurately identifying arecanut plant diseases. This breakthrough, while achieved with ResNet, opens avenues for further improvement by integrating other architectures such as MobileNet. This study represents a significant advancement in automated disease detection in agriculture, promising enhanced crop management and yields for arecanut farmers.