Abstract: Arecanut, commonly known as betel nut, is a significant tropical crop, with India being the second-largest producer and consumer worldwide. Throughout its lifecycle, it faces various diseases, affecting its roots, leaves, and fruits. Currently, disease detection relies solely on visual observation, requiring farmers to meticulously inspect each crop periodically. This project proposes a system utilizing Convolutional Neural Networks (CNNs) to detect diseases in arecanut leaves and trunk, offering corresponding remedies. CNNs are Deep Learning algorithms designed to analyze images by assigning learnable weights and biases to different features, thereby distinguishing between them. To train and validate the CNN model, a dataset comprising healthy and diseased arecanut samples was curated. The dataset was split into training and testing sets in an 80:20 ratio. For model compilation, categorical cross-entropy was employed as the loss function, with adam serving as the optimizer function and accuracy as the metric. Training the model over 50 epochs yielded high validation and test accuracies with minimal loss. The proposed approach demonstrated effectiveness, achieving a remarkable 98% accuracy in identifying arecanut diseases.
Keywords: Arecanut, betel nut, disease detection, Convolutional Neural Networks (CNN), deep learning, image classification, dataset creation, training, validation, optimization, accuracy, remedies, agricultural technology.
|
DOI:
10.17148/IJARCCE.2024.13387
