Weed and crop identification for a weeder

Currently, many farmers rely on herbicides to control weed growth, which can have negative effects on the environment, human health, and the quality of agricultural products. By developing and producing an additional agricultural machine, the project seeks to provide farmers with a more sustainable and environmentally friendly alternative.

From press release:

The aim of the project is to develop and produce an additional agricultural machine to bring cultivation closer to organic products (especially organic vegetables) to a wider range of farmers, thereby reducing the use of herbicides in vegetable fields. The objective will be achieved with the help of the mechanical removal of weeds not only between the rows, but above all by cultivating weeds between plants within the row, at the appropriate vegetative stage of the selected crop. For identification of the target crop, a camera array will be used, followed by software analysis images using neural network models. This will significantly reduce unwanted contamination with minimal damage to the selected crops. The machinery to be developed will not be single-purpose, it will be used, for example, in the cultivation of several types of vegetables or beet fields.

Our tech partners:

• SMS CZ, s.r.o.
• Výzkumný ústav zemědělské techniky, v. v. i.
• Simple Engineering s. r. o.
• Karel Dryák
• FlowerChecker s.r.o.

Our role in the project is to develop the image recognition software to classify plant and crop species but also determine the plant positions on the picture and control the weeder accordingly.