Title :
Partial least square regression for quantitative evaluation of small anomalies in non-destructive testing
Author :
Le Bihan, Yann ; Marchand, Claude ; Pávó, József ; Krebs, Guillaume
Author_Institution :
Lab. de Genie Electr. de Paris (LGEP), Univ. Pierre et Marie Curie, Gif-sur-Yvette, France
Abstract :
Partial least square regression is applied for the processing of eddy current testing data used for the reconstruction of the shape parameters of cracks. The procedure is applied for cracks having very small spatial extensions. It is found that the reconstruction of the area of such cracks is considerably more robust than the reconstruction of their length and depth parameters.
Keywords :
cracks; eddy current testing; least squares approximations; regression analysis; area reconstruction; crack shape parameters; depth parameters; eddy current testing; length reconstruction; nondestructive testing; partial least square regression; Eddy current testing; Feature extraction; Least squares methods; Nondestructive testing; Numerical simulation; Principal component analysis; Probes; Robustness; Shape; Spatial databases;
Conference_Titel :
Electromagnetic Field Computation (CEFC), 2010 14th Biennial IEEE Conference on
Conference_Location :
Chicago, IL
Print_ISBN :
978-1-4244-7059-4
DOI :
10.1109/CEFC.2010.5481758
