Purchase instant access (PDF download and unlimited online access):
Winter wheat fungal diseases, responsible for high yield losses, can be assessed by computer vision to increase phenotyping performance. This study aims to compare multispectral imagery based on remote and proximal sensing for disease detection. Wavelength selection was achieved by ANOVA and stepwise regression. Prediction of disease severity was performed by means of an artificial neural network based on proximal sensing data. Septoria tritici blotch (STB) requires proximal measurements, but stripe and brown rusts can be detected from UAVs and from the ground. Prediction results obtained gave R2 of 0.55 and 0.57 for STB and stripe rust respectively.