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

Sophian, Ali; Tian, Gui Yun; Taylor, David J.; Rudlin, John

doi:10.1016/s0963-8695(02)00069-5

Cited by 294 publications

(135 citation statements)

References 7 publications

Supporting

Mentioning

130

Contrasting

Unclassified

Order By: Relevance

“…The most popular and widely used feature extraction approach is Principal Component Analysis (PCA) [25] introduced by Karl. Principal component analysis (PCA) consists into an orthogonal transformation to convert samples belonging to correlated variables into samples of linearly uncorrelated variables.…”

Section: Feature Extraction/transformationmentioning

confidence: 99%

A Review on Dimensionality Reduction Techniques

Jindal¹,

Kumar²

2017

IJCA

View full text Add to dashboard Cite

Progress in digital data acquisition and storage technology has resulted in exponential growth in high dimensional data. Removing redundant and irrelevant features from this highdimensional data helps in improving mining performance and comprehensibility and increasing learning accuracy. Feature selection and feature extraction techniques as a preprocessing step are used for reducing data dimensionality. This paper analyses some existing popular feature selection and feature extraction techniques and addresses benefits and challenges of these algorithms which would be beneficial for beginners..

show abstract

Section: Feature Extraction/transformationmentioning

confidence: 99%

A Review on Dimensionality Reduction Techniques

Jindal¹,

Kumar²

2017

IJCA

View full text Add to dashboard Cite

show abstract

“…It has numerous promising merits [8][9][10][11] such as: Contactless, high sensitivity and rapid inspection over a large region.…”

Section: Introductionmentioning

confidence: 99%

Pattern Deep Region Learning for Crack Detection in Thermography Diagnosis System

Gao

et al. 2018

Metals

View full text Add to dashboard Cite

Eddy Current Pulsed Thermography is a crucial non-destructive testing technology which has a rapidly increasing range of applications for crack detection on metals. Although the unsupervised learning method has been widely adopted in thermal sequences processing, the research on supervised learning in crack detection remains unexplored. In this paper, we propose an end-to-end pattern, deep region learning structure to achieve precise crack detection and localization. The proposed structure integrates both time and spatial pattern mining for crack information with a deep region convolution neural network. Experiments on both artificial and natural cracks have shown attractive performance and verified the efficacy of the proposed structure.

show abstract

“…Specific to eddy current testing, PCA has been applied to steam generator tube signals to facilitate their interpretation [11], [12]. PCA has also been used with pulsed eddy current, where it was shown to enhance the classification of defects [13], and to detect defects in multilayer aluminum lap joints [14], [15] and aircraft structures [16], [17]. In this work, principal components analysis was performed on the Desjardins et al plate model of the multifrequency PT to CT gap measurement [7], [6].…”

Section: Introductionmentioning

confidence: 99%

Comparison of analytical eddy current models using principal components analysis

Contant¹,

Luloff²,

Morelli³

et al. 2017

AIP Conference Proceedings

View full text Add to dashboard Cite

Abstract.Monitoring the gap between the pressure tube (PT) and the calandria tube (CT) in CANDU ® fuel channels is essential, as contact between the two tubes can lead to delayed hydride cracking of the pressure tube. Multifrequency transmit-receive eddy current non-destructive evaluation is used to determine this gap, as this method has different depths of penetration and variable sensitivity to noise, unlike single frequency eddy current non-destructive evaluation. An Analytical model based on the Dodd and Deeds solutions, and a second model that accounts for normal and lossy selfinductances, and a non-coaxial pickup coil, are examined for representing the response of an eddy current transmit-receive probe when considering factors that affect the gap response, such as pressure tube wall thickness and pressure tube resistivity. The multifrequency model data was analyzed using principal components analysis (PCA), a statistical method used to reduce the data set into a data set of fewer variables. The results of the PCA of the analytical models were then compared to PCA performed on a previously obtained experimental data set. The models gave similar results under variable PT wall thickness conditions, but the non-coaxial coil model, which accounts for self-inductive losses, performed significantly better than the Dodd and Deeds model under variable resistivity conditions.

show abstract

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

Cited by 294 publications

References 7 publications

A Review on Dimensionality Reduction Techniques

A Review on Dimensionality Reduction Techniques

Pattern Deep Region Learning for Crack Detection in Thermography Diagnosis System

Comparison of analytical eddy current models using principal components analysis

Contact Info

Product

Resources

About