Mobile-Based Deep Learning for Yam Disease Diagnosis

Abstract

In many nations, tuber crops are crucial for both food security and the economy, constituting the staple diet for the Masses. Diseases make crops deviate from their normal growth, reducing crop yield and production. West Africa's yam productivity (yield) declined by 18.74% between 2009 to 2019, while productivity (yield) in Nigeria decreased by 23.47% between 2009 and 2019. Hence the need to build an intelligent system to assist crop growers to improve yields. The Convolution Neural Networks (CNN) deep learning model was used in this study to develop an intelligent mobile-based system for detecting Yam diseases. It was shown using a JAVA/XML Graphical User Interface (GUI). Three disease categories, namely Yam Anthracnose, Yam Mosaic Virus and Healthy were used in this study. The test data's total accuracy was 81.7%. Yam growers can utilize the GUI application program because it was designed to be user-friendly.