Fight Black-Grass using Drones and Artificial Intelligence (AI)

Fight Black-Grass using Drones and Artificial Intelligence (AI)

Project status

Ongoing

Project Manager

Category/Area

Research in Systems Engineering

Invasive species represent one of the greatest challenges for global agriculture, the challenge lies in the ability of invasive species to quickly establish themselves, change their expression and spread in new environments. The new species can cause significant damage to agricultural land, reduce harvests and increase the costs of control and management.

The effects of invasive species range from competing with native plants for resources to introducing new diseases and providing scope for pests that can negatively affect both plant and wildlife.

Against this background, new technological projects and innovations have become crucial to developing sustainable pest management practices that ultimately aim to reduce dependence on chemical pesticides and to tackle invasive species in new and more effectively manner. Through the use of new advanced technology, new conditions and methodology for precision agriculture are created. This project aims to create by studying how weeds can be precisely detected and also combated using drones and Artificial intelligence (AI).

The project not only aims to control invasive species but also promotes a future more sustainable and environmentally friendly agriculture. By integrating innovative technologies and methodologies, agriculture can reduce its environmental impact, improve food safety and ensure a more sustainable future for farming.

Facts

Duration

2023 - 2024

Budget

3,1 Mkr

Contact Person

Mattias

Mattias Dahl

mattias.dahl@bth.se

Publications

Effects of Foreground Augmentations in Synthetic Training Data on the Use of UAVs for Weed Detection
S Hallösta, MI Pettersson, M Dahl, Northern Lights Deep Learning Conference 2024. NLDL 2024. Tromsø, Norway, Jan 09 2024, https://www.nldl.org/

Multispectral Image Registration and Sensor Calibration for Low-Altitude Agricultural Drones, Simon Hallösta, Saleh Javadi, Mattias Dahl, and Mats I. Pettersson, IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2024), Athens, Greece, July 2024.

In order to ”learn” AI function how Black-grass looks, the researchers have loaded the feature with a large number of images of the invasive grass.

To find the Black-grass, a drone carrying a camera is used.

The GPS coordinates are sent to agricultural tractors, which then spray these spots precisely, instead of spraying the entire field.

Participants

Mattias Dahl

Mattias Dahl
Professor

Saleh Javadi

Saleh Javadi
Biträdande universitetslektor

Simon  Hallösta

Simon Hallösta
PhD Student