VGG16-Based Deep Learning Approach for Accurate Detection of Fruit Diseases in Precision Agriculture

Abstract

Early and accurate disease detection in fruit is crucial for precision agriculture to improve yields and provide healthy crops. This research proposes a deep learning (DL) method based on VGG16 to detect fruit illnesses from high-resolution images. The method uses VGG16 architecture, which has been fine-tuned using massive image datasets, to detect fruit illnesses. The proposed technique takes images of fruits that have several prevalent illnesses and uses those images to extract deep characteristics that allow for accurate classification. This system accurately identifies illnesses, including apple scab, citrus greening, and bacterial spots, by combining a strong convolutional neural network (CNN). The dataset used to train and test the model to simulate real-world situations contains images of healthy and sick apples taken under different lighting circumstances. The evaluation of performance indicators, including recall, accuracy, precision, and F1-score, shows the model's efficiency and accuracy in illness detection. Reduced pesticide consumption and optimized resource allocation result from this automated detection system's ability to allow farmers to monitor crop health and make prompt choices continually. It provides a useful and scalable tool for managing diseases in modern precision agriculture. A DL model based on VGG16 was created for accurate fruit disease diagnosis in precision agriculture, with an accuracy of 96.5%, allowing early intervention to improve crop health and production sustainability.