Master's thesis defense by Maria Ovchinnikova – Niels Bohr Institute - University of Copenhagen

Niels Bohr Institute > Calendar > NBI Calendar 2017 > Master's thesis defens...

Master's thesis defense by Maria Ovchinnikova

Title: Automated Wind Turbine Blade Inspection

In this thesis, an effective method for automated defect detection on wind turbine blades is proposed. Over a period of time the wind power industry has collected large imagery data of wind turbine blades. Today, a human operator systematically assesses all images captured, maps and reports the defects found in accordance with a defect catalogue. Based on the fact that most defects can be regarded as combinations of several image features, the proposed method first extracts all the features as an input vector and then recognizes the defects using a trained classifier. The proposed method is trained and evaluated using a constrained defect dataset of 50 000 images. As a result, the method reached state-of-the-art 99% accuracy in recognition rates. This shows the superiority of the proposed method for wind turbine blade defect recognition.

Please note that the data used in this thesis are strictly confidential. In other words, the defense is closed, not public.