Automatic Defect Identification of Eddy Current Pulsed Thermography Using Single Channel Blind Source Separation

Pulsed inductive thermal wave radar (PI-TWR) using cross correlation matched filtering in eddy current thermography

“…The principle and diagram of ECPT were shown in detail in previous works [24,25,36]. The excitation signal is a small period of high frequency current.…”

“…The results from ECLT and ECPPT have illustrated that the non-uniform heating effect can be eliminated and defect detectability can be significantly improved through phasegram or phase information. In addition, several signal processing and blind source separation methods, such as principal components analysis (PCA), independent components analysis (ICA), were used to improve the defect detectability [9,[22][23][24][25] dynamic range and resolution for detection depth are still not satisfactory.…”

Thermography spatial-transient-stage mathematical tensor construction and material property variation track

“…where electric conductivity s is dependent on temperature and s 0 is the conductivity at the reference temperature T 0 and a is the temperature coefficient of resistivity, which describes how resistivity varies with temperature [18]. In general, by taking account of heat diffusion and Joule heating, the heat conduction equation of a specimen can be expressed as:…”

“…Cheng et al evaluated notches in carbon fiber reinforced plastic material through analysis of the surface heating pattern [17]. Gao et al proposed source separation algorithm on ECPT for automatic crack detection and identification [18]. All of the above cited researches recognize that the basic physical mechanism corresponding to the general behavior of ECPT is the result of Joule heating via eddy current and heat diffusion [19].…”

Pattern Deep Region Learning for Crack Detection in Thermography Diagnosis System

“…To achieve automatic crack detection and identification for the experimental data from the ECPT system, a blind source separation algorithm was reported [20]. Methods based on sparse decomposition exhibited their robustness for both man-made specimens and samples with natural defects [21][22][23][24]. These methods assume that regions with defects are areas with the highest sparsity, while the low-rank matrix, which is considered as background, is separated to extract sparse components.…”