Download PDFOpen PDF in browserWheat Fusarium Head Blight Disease Severity Estimation Using UAS Multispectral Imagery and Machine LearningEasyChair Preprint 152336 pages•Date: October 18, 2024AbstractWheat is an important primary crop that feeds billions of people worldwide. Wheat diseases, particularly Fusarium Head Blight (FHB) disease often have severe effects on wheat yield quantity and quality, and also potentially threaten the health of humans and livestock. Traditional field surveying-based methods for monitoring and assessment of wheat diseases are time-consuming and inefficient. Remote sensing approach, particularly aerial imaging using Uncrewed Aircraft Systems (UAS) has become an essential tool for fine-scale and rapid field scouting and crop disease monitoring in recent years. The work is to investigate the potential of combination of high-resolution UAS multispectral imagery with machine learning (ML) methods in detection of FHB disease severity. Two experimental wheat fields are setup in Brookings, South Dakota, USA in 2022. The severity of FHB disease was assessed periodically through visual observation; and synchronous UAS flights were conducted to collect multispectral imagery. UAS multispectral imagery-based canopy spectral and texture features were derived, and used as input variables to develop ML-based classification and regression models for FHB disease severity estimations. Conventional ML methods such as Support Vector Machine (SVM), Random Forest (RF), along with deep learning models, Deep Neural Network (DNN), and One-Dimensional Convolutional Neural Network (1DCNN) were employed. The results show that both canopy spectral and texture features are important variables for wheat FHB disease severity estimations. Additionally, UAS remote sensing, coupled with ML-based classification and regression approaches is a viable approach to rapidly and accurately estimate and assess wheat FHB disease severity at a fine-scale for large areas Keyphrases: Blight (FHB), Fusarium Head, Spectral features, Uncrewed Aircraft Systems (UAS), machine learning, texture features
|