Detection of coating defects in high gas turbine blades by image processing

Turbine blades, located after the combustion chamber in a gas turbine engine, are in a critical position and exposed to high temperature and pressure. Coating degradation is a frequentative phenomenon which can lead to blade failure, so the turbine blade needs checking in specific intervals by visual inspection and borescope. In this article, a methodology is developed by image processing (IP) techniques to measure the degraded coating area. In this study, discoloration in coating as an introductory indication of degradation has been studied and diagnosed by IP, while a formula, containing the combination of these defects, has been set as the final criteria for automated decision making. Results have illustrated an acceptable confirmation between computer vision and experimental approaches, which can be utilized in borescope image processing.