ASABE Technical Library - Abstract

Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, view access and subscription options

Advanced AI Framework for Phenotyping of Fusarium Head Blight Disease in Wheat

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  2026 ASABE Annual International Meeting  2601083.(doi:10.13031/aim.202601083)
Authors:   Karishma Kumari, Ali Mirzakhani Nafchi
Keywords:   FHB; DL; Attention U-Net; YOLOv11; Disease Detection and Severity Estimation; Precision Agriculture

Abstract. Fusarium Head Blight (FHB) is a significant fungal disease in wheat and barley that degrades grain quality and contaminates crops with toxic mycotoxins. Traditional disease assessment is labor-intensive, subjective, and time-consuming, which limits its usefulness in plant breeding efforts. This work describes a field-based digital phenotyping method for automated FHB identification and severity evaluation, which employs a custom-built imaging platform to capture multi-angle wheat head pictures. Roboflow was used to create ground-truth annotations, and illness analysis was carried out using two deep learning models: Attention U-Net for segmentation and YOLOv11 for spike detection and classification. An interactive Flask-based web dashboard allows users to input pictures, view severity maps, and produce CSV or PDF reports in near real-time. The suggested method offers a scalable and economical alternative to manual disease scoring, allowing for crop health monitoring and data-driven selection of disease-resistant types.

(Download PDF)    (Export to EndNotes)