Detecting Apple and Maize Plant Diseases using GAN and GNN

The sustenance of the global demand for food supply is highly dependent on the management practices and technology enhances agricultural performance by increasing both yield and quality over time. The advancements of agricultural techniques aimed at eliminating food security concerns have been pegged on the ability to detect and diagnose various plant pathogens. It is known that certain diseases in crops for instance, apple and maize diseases, are capable of reducing yield hence leading to food inadequacy and loosing markets to competitors for the farmers. This reminds us, that automated systems for diagnosis and detection of plant diseases should be developed to take advantage of healthy machine learning algorithms and advancements in technology on various datasets. In this paper, we propose a novel approach that uses state-of-the-art deep learning models such as GANs, Resnet 152, and Graph Neural Networks that can be used in the diagnosis and treatment of plant diseases. Use of GANs helps in increasing the diversity of the dataset by augmentation, extraction of linear hierarchical features is performed by Resnet-152 whereas GNNs are utilized to represent the data and model the relationships. The integration of these approaches boosts the efficiency of the computed disease detection systems by providing relevant information to the farmers. Therefore, contributing to the implementation of precision agriculture by facilitating early and accurate disease prediction, the system minimizes the use of chemicals such as pesticides and herbicides that may pollute the environment and promotes healthy farming. Also, it allows farmers to manage the situation effectively to avoid the spread of the disease thereby improving the quality of the crops and increasing the yield. It is observable that the usage of this hybrid model has the potentials the transform agricultural practices in this modern age of technology to effectively address food sufficiency concerns.