Quantum Resnet18 for Classifying and Predicting the Maize Leaf Diseases

The agriculture industry contributes one-third of the worldwide Gross Domestic Product (GDP). Moreover, a substantial number of developing countries rely on their agricultural output as it offers job prospects for a considerable portion of the impoverished population. This necessitates the use of techniques to guarantee the precise and effective identification of plant diseases to minimize any detrimental consequences on the crops. Recently, the application of quantum-enhanced techniques to machine learning tasks including recommendation systems, classification, and regression has increased significantly. For this reason, this work suggests using the hybrid quantum-based Resnet18 transfer learning model to categorize and diagnose diseases that impact the leaves of maize plants. The dataset employed consists of four categories including Leaf Blight, Grey Leaf spots, Common Rust, and Healthy leaves. When compared to Resnet18 without quantum, this method outperforms it in terms of recall, accuracy, precision, and f1-score. The validation accuracy of 87% is achieved using the model. In the future, the proposed effort can serve as a practical tool to aid farmers in recognizing the diseases and safeguarding their maize crops.