Crop protection products (i.e. pesticides) play a strategic role in securing worldwide food production. Nevertheless, major societal concerns are raised about risks for human health and environment which are being addressed by policy actions for reducing the dependence of agriculture on pesticides.
A primary contribution to this goal is expected to come from precision crop protection. In opposition to current uniform field treatments, precision protection is founded on the spatial modulation of treatments, tailored to site/time-specific needs of protecting the crop from pest pressure and expected spreading.
To be implemented, precision protection requires the detection of early stages of pest infections, as this enables selective treatments of initial foci, preventing epidemic spread to the whole field. Aim of this project is to combine advanced crop sensing technologies with state-of-the-art deep learning models in order to create the building blocks of an intelligent precision crop protection system.
Aims of the project
- Develop data fusion approaches to extract augmented information from complementary sensors
- Explore different structures of convolutional neural networks for early detection of pest infection from elaborated data
- Field path optimization of automated platforms in search of early infections.
The PRECISION project is funded by Università degli studi di Milano LA STATALE, under the call SEED 2019 – Bando Straordinario per Progetti Interdipartimentali. PRECISION project belongs to the ensemble of 42 projects funded by the SEED 2019 call.
- Precision Agriculture
- Huang et al, A fully convolutional network for weed mapping of unmanned aerial vehicle (UAV) imagery
- Mohanty et al, Using deep learning for image-based plant disease detection
- Oberti et al, Automatic detection of powdery mildew on grapevine leaves by image analysis: Optimal view-angle range to increase the sensitivity
- Convolutional Neural Networks
- Deep Learning Architectures (book)
- Related European Projects