Poster # 510
Ali M Nafchi 1, Karishma Kumari 1, Ahmed Abdalla 1, Karl Glover 1, Julie Thomas 1, Sunish Sehgal 1, and Kwanghee Won 2
1. Department of Agronomy, Horticulture, and Plant Science, South Dakota State University, Brookings, SD, 57007
2. McCombish Department of Electrical Engineering and Computer Science, South Dakota State University, Brookings, SD 57007
Corresponding author: Ali Nafchi, ali.nafchi@sdstate.edu
Nafchi, Ali M
Fusarium head blight (FHB) severely constrains wheat and barley breeding due to the time-, labor-, and expertise-intensive nature of field scoring. We developed and field-tested an innovative 360° Deep Scanning Robot that leverages advanced imaging, sensor fusion, and AI-driven analytics to detect FHB and quantify disease severity in real time. Preliminary deployments by the Precision Agriculture team at SDSU produced highly encouraging results, generating cardiograph-like graphical outputs that deliver objective, line-by-line ratings across breeding plots. We have proposed a next-level, automated Head-by-Head FHB scoring system featuring an adaptive, height-adjustable scanning mechanism for plant-head-level assessment for variable canopy heights. The system integrates trained computer-vision models with high-throughput field phenotyping to provide precise, rapid (<10 min) and reproducible resistance ratings. System integration, real-time algorithm optimization, and validation across multi-location field trials by benchmarking against expert manual assessments and stress-testing performance across genotypes, crop heights, and environments. Deliverables include an operational prototype, validated accuracy and throughput metrics, and user-ready training resources and best-practice guides. This project directly aligns with USWSI research priorities and, once fully optimized, promises scalable, and efficient FHB screening to accelerate genetic gain in wheat and barley breeding programs.
© Copyright
2025 by individual authors. All rights reserved. No part of this abstract or paper publication may be reproduced without prior permission from the applicable author(s).