IoT-based Agriculture: Deep Learning in Detecting Apple Fruit Diseases

Abstract IoT-based agricultural environment provides the potential of using different technologies to help agricultural industry in monitoring and preventing fruit diseases. In IoT-based agriculture, machine learning and image processing techniques are the main expertise's required to propose and develop efficient method for diagnosing and preventing infection in agricultural products. In this article, we propose a method to detect infections in apple fruit and timely prevention of further infections caused by environmental factors. Deep learning, which proved its capability in image processing and classification, is used to classify apple images. Deep neural network with different convolution layers and different number of neurons are examined and evaluated. The results are evaluated in terms of accuracy, sensitivity, specificity and ROC curve. In addition, a comparison has been made with the research reported on apple image classification and the superiority of the proposed method is shown.