Case study card:
Automated crop monitoring and defects detection
Primary Industries and Infrastructure
We were challenged to detect sugarbeet weeds at an early stage based on close-up images of single plants grown under greenhouse conditions.
A mobile app was developed that Agroservice employees and farmers could use to identify weeds in the field using smartphone cameras. The application connects to the Artificial Intelligence and Computer Vision API where deep learning algorithms perform the analysis of particular pictures. Our models achieved above 90% of accuracy.
Automated growth stage detection as well as automated categorization of sugar beets.
— Python, Docker, PyTorch, CNN