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

Autonomous Field Phenotyping for Rapid Disease Screening Using AI and Multi-View Imaging

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

Citation:  2026 ASABE Annual International Meeting  2600843.(doi:10.13031/aim.202600843)
Authors:   Karishma Kumari, Ali Mirzakhani Nafchi
Keywords:   FHB, Deep Learning, 360-PICS system, Plant morphology, Disease management, Deep learning and Artificial Intelligence

Abstract. One of the most damaging diseases affecting wheat and barley production globally is Fusarium Head Blight (FHB), which results in significant yield reduction, deterioration of grain quality, and mycotoxin contamination. Conventional illness assessment techniques are subjective, time-consuming, and inappropriate for large-scale phenotyping applications. A Head-by-Head Artificial Intelligence (AI) framework combined with a 360-PICS system for automated FHB scoring in field settings is presented in this study. To increase the effectiveness of illness diagnosis, the suggested approach integrates deep learning-based segmentation and localization methods with close-range omnidirectional imaging. After evaluating several convolutional neural network (CNN) architectures, Attention U-Net showed the highest overall performance for detecting infected spike locations under challenging field conditions. The system designed enables crop breeding and precision agricultural applications and offers a scalable, high-throughput substitute for traditional visual scoring techniques.

(Download PDF)    (Export to EndNotes)