A feature extraction technique based on principal component analysis for pulsed Eddy current NDT

Pulsed Eddy current (PEC) is a new emerging NDT technique for sub-surface defect detection. The technique mainly uses the response peak value and arrival to detect and quantify the defects. This could suffer from noise and be not sufficient to extract more information about the defects, e.g. depth information of defects. This paper introduces the application of principal component analysis in extracting information from PEC responses. A comparative test carried out shows that the introduced technique has performed better than the conventional technique in the classification of defects