Leaf Disease Identification on Infrared Images based on ConvNext v2 model

Identifying diseases in leaves plays a crucial role in ensuring crop productivity and plant health. In this work, we introduce an innovative approach utilizing infrared images for leaf disease identification, employing a ConvNeXt v2-based model. Infrared imaging allows for better insights into plant health as it can detect plant temperature differences, which usually indicate disease. We take a version of the ConvNext v2 model known for its better ability to extract features and classify and fine-tune them for the task. This method outperforms conventional methods and shows optimum accuracy in detecting different diseases of a leaf. The research method is a significant breakthrough in automated plant disease diagnostics and contributes to sustainable agriculture through timely interventions.