Efficient feature selection for neural network based detection of flaws in steel welded joints using ultrasound testing

HIGHLIGHTSA comparative study on different feature extraction techniques is presented.PCA is used to provide a compact set of representative features.Statistical significance tests prove to be an interesting alternative to PCA.Different neural network based classifier designs are evaluated. ABSTRACT This work studies methods for efficient extraction and selection of features in the context of a decision support system based on neural networks. The data comes from ultrasonic testing of steel welded joints, in which are found three types of flaws. The discrete Fourier, wavelet and cosine transforms are applied for feature extraction. Statistical techniques such as principal component analysis and the Wilcoxon‐Mann‐Whitney test are used for optimal feature selection. Two different artificial neural network architectures are used for automatic classification. Through the proposed approach, it is achieved a high discrimination efficiency by using only 20 features to feed the classifier, instead of the original 2500 A‐scan sample points.

[1]  Xiaofeng Li,et al.  Oil Spill Mapping and Measurement in the Gulf of Mexico With Textural Classifier Neural Network Algorithm (TCNNA) , 2013, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[2]  A. Masnata,et al.  Neural network classification of flaws detected by ultrasonic means , 1996 .

[3]  Masoud Nikravesh,et al.  Feature Extraction - Foundations and Applications , 2006, Feature Extraction.

[4]  Luminita Moraru,et al.  Cardiac Cycle Phase Estimation in 2-D Echocardiographic Images Using an Artificial Neural Network , 2013, IEEE Transactions on Biomedical Engineering.

[5]  Robert W. Messler,et al.  Joining of Materials and Structures: From Pragmatic Process to Enabling Technology , 2004 .

[6]  Soo-Chang Pei,et al.  Two-Dimensional Orthogonal DCT Expansion in Trapezoid and Triangular Blocks and Modified JPEG Image Compression , 2013, IEEE Transactions on Image Processing.

[7]  Ivan Bartoli,et al.  Defect Classification in Pipes by Neural Networks Using Multiple Guided Ultrasonic Wave Features Extracted After Wavelet Processing , 2005 .

[8]  Ian T. Jolliffe,et al.  Principal Component Analysis , 2002, International Encyclopedia of Statistical Science.

[9]  Silvia Figini,et al.  Computational Statistics and Data Analysis Making Classifier Performance Comparisons When Roc Curves Intersect , 2022 .

[10]  Hao Yu,et al.  Improved Computation for Levenberg–Marquardt Training , 2010, IEEE Transactions on Neural Networks.

[11]  Romesh Nagarajah,et al.  Feature selection for neural network based defect classification of ceramic components using high frequency ultrasound. , 2015, Ultrasonics.

[12]  R. Lyman Ott.,et al.  An introduction to statistical methods and data analysis , 1977 .

[13]  Erkki Oja,et al.  Independent Component Analysis , 2001 .

[14]  Harry L. Van Trees,et al.  Detection, Estimation, and Modulation Theory, Part I , 1968 .

[15]  M. Melamed.  Detection , 2021, SETI: Astronomy as a Contact Sport.

[16]  José Manoel de Seixas,et al.  Modified post-nonlinear ICA model for online neural discrimination , 2010, Neurocomputing.

[17]  C Cerrillo,et al.  New contributions to granite characterization by ultrasonic testing. , 2014, Ultrasonics.

[18]  Eduardo F. Simas Filho,et al.  Unsupervised statistical learning applied to experimental high-energy physics and related areas , 2016 .

[19]  S. Seyedtabaii.  Performance Evaluation of Neural Network Based Pulse-Echo Weld Defect Classifiers , 2012 .

[20]  Giles M. Foody,et al.  Feature Selection for Classification of Hyperspectral Data by SVM , 2010, IEEE Transactions on Geoscience and Remote Sensing.

[21]  Rui Zhong,et al.  Dictionary based surveillance image compression , 2015, J. Vis. Commun. Image Represent..

[22]  Barbara Pfeffer,et al.  Fundamentals And Applications Of Ultrasonic Waves , 2016 .

[23]  L.A.L. de Almeida,et al.  Vibration Monitoring of On-Load Tap Changers Using a Genetic Algorithm , 2005, 2005 IEEE Instrumentationand Measurement Technology Conference Proceedings.

[24]  David Poole,et al.  Linear Algebra: A Modern Introduction , 2002 .

[25]  João M.A. Rebello,et al.  The use of artificial neural network in the classification of pulse-echo and TOFD ultra-sonic signals , 2005 .

[26]  Rifai Chai,et al.  Brain–Computer Interface Classifier for Wheelchair Commands Using Neural Network With Fuzzy Particle Swarm Optimization , 2014, IEEE Journal of Biomedical and Health Informatics.

[27]  S. Sambath,et al.  Automatic Defect Classification in Ultrasonic NDT Using Artificial Intelligence , 2011 .

[28]  Maria C. S. Albuquerque,et al.  Decision support system for ultrasound inspection of fiber metal laminates using statistical signal processing and neural networks. , 2013, Ultrasonics.

[29]  Francisco Herrera,et al.  Evaluating the classifier behavior with noisy data considering performance and robustness: The Equalized Loss of Accuracy measure , 2016, Neurocomputing.

[30]  Stéphane Mallat,et al.  A Wavelet Tour of Signal Processing - The Sparse Way, 3rd Edition , 2008 .

[31]  Simon Haykin,et al.  Neural Networks and Learning Machines , 2010 .

[32]  Eduardo A. B. da Silva,et al.  Digital Signal Processing System Analysis and Design: Preface , 2010 .