Banana Fruit Disease Detection and Categorization Utilizing Graph Convolution Neural Network (GCNN)

Agriculture seem to be the primary sector of income for the majority of the world’s population, whether there are issues in the primary sector, it would harm the population’s livelihood. Nowadays, fruits are susceptible to various illnesses, also manual disease diagnosis in fruit is quite difficult in the early stages. In which, banana is a fruit that grows with a very high yield in several Asian countries and is one of the most profitable fruits on the market. However, due to a variety of factors, the fruits become infected with a variety of diseases, which destroy the entire fruit and result in much lower product output. Thus, this study provides deep learning architecture in which a novel graph convolutional neural network is introduced to overcome the main difficulties with phytopathology. Moreover, this neural network deals with banana fruit disease detection and categorization. As a result, the proposed neural network classifies the types of disease in the fruit image with improved classification accuracy and the simulation results proved that the proposed neural network gives better result compared to other neural networks.