Abstract
Plant diseases significantly impact agricultural yields and food security worldwide. Accurate and timely disease detection is essential for effective crop management. This research explores an advanced approach to plant leaf disease detection, focusing on corn or maize leaves, through the integration of Specim IQ hyperspectral imaging and Residual Neural Networks (ResNet-50 and ResNet-101). The Specim IQ sensor is employed for data acquisition, providing high-resolution hyperspectral imagery. The methodology involves preprocessing techniques tailored to the Specim IQ sensor, customization of ResNet architectures for hyperspectral data, and comprehensive training and validation processes. This research involves dataset acquisition, pre-processing, and training the models with spectral information. Results indicate a notable improvement in disease detection accuracy, showcasing the potential of ResNet architectures and spectral imaging in plant pathology. The study evaluates the model performance using metrics such as accuracy, precision, recall, and F1 score. A specific case study on corn or maize leaves unveils insights into the potential of hyperspectral imaging for precise disease identification. This research contributes to the advancement of plant disease detection methodologies, showcasing the synergy between cutting-edge hyperspectral imaging technology and powerful deep learning architectures.
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright (c) 2024 African Journal of Biomedical Research