Abstract
Composite materials, such as Carbon-Fiber-Reinforced Plastic (CFRP), are used in many industries because they have advantages over more traditional materials. However, CFRPs may have structural flaws, because mechanical stress or manufacturing process, that represent an important risk for the safe operation of CFRP-made structures. This study analyzes the performance in detection of internal defects, by means of training and operating Support Vector Machines (SVM) with thermal contrast information obtained from Background Thermal Compensation by Filtering (BTCF) technique. IR images were obtained by using an Active Pulsed Thermography (PT) system, under two different conditions, for inspection of a 20 × 20 cm CFRP slab with 25 squared Teflon insertions as emulated defects. Detection results show that the combination of BTCF contrast technique and SVM classifier leads to a greater sensibility (22 of 23 defects considered) than other combinations of thermal contrast, feature selection and classifiers proposed in previous works.
Similar content being viewed by others
References
Ibarra-Castanedo, C., et al.: Comparative study of active thermography techniques for the nondestructive evaluation of honeycomb structures. Res. Nondestruct. Eval. 20(1), 1–31 (2009)
Ibarra-Castanedo, C., Tarpani, J., Maldague, X.: Nondestructive testing with thermography. Eur. J. Phys. 34(6), 91–109 (2013)
Valdés, M., Inamura, M., Valera, J., Yao, L.: Multidimensional filtering approaches for pre-processing thermal images. Multidimens. Syst. Signal Process. 17(4), 299–325 (2006)
Alsaadawi, Z., Netzelmann, U.,: Spatio-temporal filtering of active thermography data for noise reduction and data compression. In Proceedings of 10th International Conference on Quantitative InfraRed Thermography (QIRT 2010). Québec, Canada (2010)
Restrepo-Girón, A., Loaiza-Correa, H.: New advances in multidimensional processing for thermal image quality enhancement. In: Santhi, V. (ed.) Recent advances in applied thermal imaging for industrial applications, pp. 202–248. IGI Global, Hershey PA (2017)
García, D.: Robust smoothing of gridded data in one and higher dimensions with missing values. Comput. Stat. Data Anal. 54(4), 1167–1178 (2010)
Benítez, H., Loaiza, H., Caicedo, E.,: Termografía activa pulsada en inspección de materiales. Técnicas avanzadas de procesado [Active pulsed thermography for materials inspection. Advanced processing techniques]. Universidad del Valle—Pontificia Universidad Javeriana. Cali, Colombia (2011)
Madruga, F., Ibarra-Castanedo, C., Conde, O., López-Higuera, J., Maldague, X.: Infrared thermography processing based on higher-order statistics. NDT&E Intern. 43(8), 661–666 (2010)
Roche, J., Leroy, F., Balageas, D.: Images of TSR coefficients: a simple way for rapid and efficient detection of discontinuities. Mater. Eval. 72(1), 73–82 (2013)
Bishop, C.M.: Pattern recognition and machine learning. Springer, Cambridge (2006)
Peeters, J., Ibarra-Castanedo, C., Sfarra, S., Maldague, X., Dirckx, J., Steenackers, G.: Robust quantitative depth estimation on CFRP samples using active thermography inspection and numerical simulation updating. NDT&E Intern. 87, 119–123 (2017)
Junyan, L., Qingju, T., Xun, L., Yang, Y.: Research on the quantitative analysis of subsurface defects for non-destructive testing by lock-in thermography. NDT&E Intern. 45(1), 104–110 (2012)
Benítez, H., Loaiza, H., Caicedo, E., Ibarra-Castanedo, C., Bendada, A., Maldague, X.: Defect characterization in infrared non-destructive testing with learning machines. NDT&E Intern. 42(7), 630–643 (2009)
Darabi, A., Maldague, X.: Neural network based defect detection and depth estimation in TNDE. NDT&E Intern. 35(3), 165–175 (2002)
Luo, Y., Yan, Z., Wang, K., Wang, L.,: Infrared Image Registration of Damage in the Aircraft Skin Based on Lie Group Machine Learning. In: 26th Chinese Control and Decision Conference (CCDC) (2014)
Dudzik, S.: Analysis of the accuracy of a neural algorithm for defect depth estimation using PCA processing from active thermography data. Infrared Phys. Technol. 56, 1–7 (2013)
Restrepo, A., Loaiza, H.: Background thermal compensation by filtering for contrast enhancement in active thermography. J. Nondestruct. Eval. 35(1), 20 (2016)
Jimenez, L., Rengifo, P.: Al interior de una máquina de soporte vectorial [Inside of a Support Vector Machine]. Rev. Ciencias Univ. del Val. 14, 73–85 (2010)
Riobó, V.: Reconocimiento de localizaciones mediante máquinas de soporte vectorial [Locations recognition by support vector macines]. Dissertation, Universidad Carlos III, Madrid (2012)
Ledezma, W.: Máquinas de Soporte Vectorial (SVM) para la Detección de Nódulos Pulmonares En Tomografía Axial Computarizada (TAC) [Support Vector Machines (SVM) for Detection of Pulmonary Nodules in Computerized Axial Tomography]. Dissertation, Pontificia Universidad Javeriana, Bogotá (2013)
Ratnayake, R., Hicks, C., Akbari, M.: A SVM based Method to Detect Color Shift Defects in IC Packages. In: APR Conference on Machine Vision Applications. Tokyo, Japan (2007)
Su, L., Shi, T., Xu, Z., Lu, X., Liao, G.: Defect inspection of flip chip solder bumps using an ultrasonic transducer. Sensors (open access). 13, 16281–16291 (2013)
Ruedas, C.: Detección Multiusuario para DS-CDMA basado en SVM [Multiuser Detection for DS-CDMA based on SVM], Dissertation, Universidad de Sevilla
Burges, C.: A tutorial on support vector machines for pattern recognition. Data Min. Knowl. Discov. 2(2), 121–167 (2002)
Acknowledgements
The authors would like to express their gratitude to the multipolar infrared vision (MIVIM) research group at the University of Laval in Quebec, Canada, for providing their facilities to produce the thermal images used in this study.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Forero-Ramírez, JC., Restrepo-Girón, AD. & Nope-Rodríguez, SE. Detection of Internal Defects in Carbon Fiber Reinforced Plastic Slabs Using Background Thermal Compensation by Filtering and Support Vector Machines. J Nondestruct Eval 38, 33 (2019). https://doi.org/10.1007/s10921-019-0569-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10921-019-0569-6