Deep Learning algorithms for the development and reduction of times in the validation of new products for crop protection (insecticides, fungicides and herbicides).
Proper and early identification of plagues in crops minimises the loss of crop performance and increases the effectiveness of treatments.
In this project, developed in conjunction with BASF, an algorithm has been created based on an in-depth neuronal network for the automatic counting of insects in their different states, enabling new products that minimise their negative impact to be validated.
As a result, we have succeeded in:
- Increasing accuracy in the effectiveness of insecticides through robust insect counting algorithms.
- Increasing the evaluation frequency through a system that captures and processes images in the field.
- Generating new knowledge through advanced statistical models.
