LodgedNet is an architecture for crop lodging prediction from UAV-acquired images of wheat and canola using a DCNN augmented with handcrafted texture features. The paper will be available soon.
The proposed architecture has been designed to work with 5-channel data: red, blue, green, red edge, and near infrared.
The main dependencies of LodgeNet are numpy, pandas, and pytorch and torchvision. See requirements.txt for the list of requirements.