Banca de QUALIFICAÇÃO: CAMILA HELENA TEIXEIRA

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
DISCENTE: CAMILA HELENA TEIXEIRA
DATA: 30/05/2025
HORA: 14:00
LOCAL: DBI02 e On-line (meet.google.com/bww-dfqd-xkd)
TÍTULO:

ENHANCING WHEAT BLAST RESISTANCE BY INTEGRATING CONVOLUTIONAL NEURAL NETWORKS ON IMAGE ANALYSIS

PALAVRAS-CHAVES:

Triticum aestivum; Pyricularia oryzae; genetic resistance, artificial intelligence; plant breeding

PÁGINAS: 25
GRANDE ÁREA: Ciências Agrárias
ÁREA: Agronomia
SUBÁREA: Fitotecnia
ESPECIALIDADE: Melhoramento Vegetal
RESUMO:

Wheat blast is one of the most important diseases affecting wheat, causing significant yield losses, particularly during the reproductive stage. Assessing the genetic resistance in the commercial germplasm is essential for both cultivar recommendation and the selection of parental lines for breeding programs. However, traditional methods for evaluating disease severity can be labor-intensive and prone to inaccuracies. In this study, we aimed to train a convolutional neural network (CNN) model to classify images of wheat spikes inoculated with four isolates of Pyricularia oryzae based on wheat blast severity. Images were captured 11 days after inoculation and preprocessed using NDVI-based segmentation and k-means clustering techniques to distinguish between healthy and symptomatic areas. Disease severity was quantified, and each spike image was classified into one of four categories: R – resistant, MR – moderately resistant, MS – moderately susceptible, and S – susceptible. CNN models were trained using the YOLOv11 nano architecture with transfer learning, allocating 75% of the dataset for training and 25% for testing. The models demonstrated high performance, with particularly strong results from the general severity and isolate-specific classification models, which achieved up to 99% accuracy. These findings highlight the effectiveness of CNNs for automated, real-time, and accurate assessment of wheat blast severity, providing valuable support in developing resistant cultivars.

MEMBROS DA BANCA:
Interno - VINICIUS QUINTAO CARNEIRO (Membro)
Externo à Instituição - ROBERTO FRITSCHE NETO - NCSU (Membro)
Externo à Instituição - MAICON NARDINO - UFV (Membro)
Externo ao Programa - JOSE MARIA VILLELA PADUA - DAG/ESAL (Suplente)
Presidente - JOSE AIRTON RODRIGUES NUNES (Membro)