MSc thesis subject: A Deep Learning Approach for Detection of Poisonous Weeds in Wide Grasslands

Unmanned aerial vehicles (UAV) shipped with on-board sensors have become an effective remote sensing (RS) tool in agriculture and environmental studies.
However, the full potential of those platforms has not been reached yet. Working with UAV it’s an added valuable point for any professional who wishes to succeed in this competitive market. Herein, you will have the opportunity to work in a novel and ambitious project using small UAV for environment monitoring and livestock/Agriculture.

Grasslands have a very positive environmental and economical impact. It promotes the production of food and helps to improve the soils health. However, there are some injurious weeds that threat this ecosystem, e.g., causing sickness in livestock. Those weeds are difficult to eradicate if not detected at an early stage.

The goal of this work is to develop a deep learning based approach for detecting poisonous weeds in high-resolution aerial imagery from a wide grassland field. This thesis will be carry out in the follow steps:

  1. Review of the state-of-the art of deep learning approaches for weed detection;
  2. Methodology design;
  3. Experiments in available dataset.


  • Andreas Kamilaris, Francesc X. Prenafeta-BoldĂș, Deep learning in agriculture: A survey, Computers and Electronics in Agriculture, Vol. 147, pp. 70-90, 2018.
  • Mads Dyrmann, Henrik Karstoft, Henrik Skov Midtiby, Plant species classification using deep convolutional neural network, Biosystems Engineering, Volume 151, pp. 72-80, 2016.


  • UAV enthusiast
  • Willing for learning novel software and hardware tools
  • Excited to work in deep learning

Theme(s): Sensing & measuring