Novel Method for Wind Turbines Blades Damage Classification using Image Processing

Abstract

Wind turbine generators are spreading around the world due to its advantages over fossil fuels. Structural monitoring of them is important to increase operation and reduce maintenance times. Visual inspection is highly influenced by the human factor due to the working conditions. Image processing supported by vision systems offers high advantages reducing times, being the software and processing algorithms, which generates added value. In this paper, a novel method for wind turbines blades damages analysis is presented using image processing and a classifier based on dimensional features. The image acquisition is performed using a reflex camera with a telephoto and geo-location enabled. The faults analyzed include cracks, edge erosion, and electric discharge.