Master Thesis

Neural point features for automatic wind turbine inspection

Work default illustration


  • If you are interested in the proposal, please contact with the supervisors.


Drones are used to capture blade pictures to periodically inspect wind turbines. The purpose is to detect blade defects, so they can be repaired in due time. This project focuses on obtaining meaningful image features that can be easily distinguishable between them, so we can later use it to stitch the images and create a panoramic/mosaic or 3D model. The goal is to develop deep-learning features inspired by classic descriptors such as SIFT or SURF. This tool would be used to locate where the picture was taken overall the wind blade. This industrial project will be executed in collaboration with the company Wind Power Lab, who will provide the image data.

Requisites: Candidates with a background in optimization, machine learning and good programming skills (Matlab/Python/C++) are particularly encouraged to apply. This project will be supervised by Raül Perez i Gonzalo and Antonio Agudo.