Deep-learning-based crack detection with applications for the structural health monitoring of gas turbines

Gas turbine maintenance requires consistent inspections of cracks and other structural anomalies. The inspections provide information regarding the overall condition of the structures and yield information for estimating structural health and repair costs. Various image processing techniques have been used in the past to address the problem of automated visual crack detection with varying degrees of success. In this work, we propose a novel crack detection framework that utilizes techniques from both classical image processing and deep learning methodologies. The main contribution of this work is demonstrating that applying filters to image data in the pre-processing phase can significantly boost the classification performance of a convolutional neural network-based model. The developed architecture outperforms compared works by yielding a 96.26% classification accuracy on a data set of cracked surface images collected from gas turbines.