Int J Performability Eng ›› 2021, Vol. 17 ›› Issue (5): 484-490.

### Fuzzy C-based Automatic Defect Detection using Barker Coded Thermal Wave Imaging

M. Muzammil Parveza,*, J. Shanmugama, and V.S. Ghalib

1. a Department of Electronics and Communication Engineering, Bharath Institute of Higher Education and Research (BIHER), Chennai, 600073, India;
b Department of Electronics and Communication Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, 522502, India
• Contact: * E-mail address: parvez190687@yahoo.co.in

Abstract: Non-Destructive Testing & Evaluation (NDT&E) is used to test the object for detection of voids/cracks that are generated due to various circumstances at the time of production and making of various materials. Among several NDT&E techniques, Non-stationary thermal wave imaging (NSTWI) is an attractive and significant method for testing those materials without changing their serviceability. NSTWI is a whole filed, non-contact and non-invasive testing modality. The present research work is focused on detecting subsurface irregularities/defects that arose during the manufacturing phase by employing a clustering based post-processing approach using Barker Coded Thermal Wave Imaging (BCTWI). The Fuzzy c-means (FCM) & K-means clustering algorithms were intended to classify and detect the defects in the test samples. Fuzzy c-means clustering enhances a better detectability compared to that of the K-means clustering by eliminating the initial counter problems that arise during detection. The obtained experimental results using Fuzzy c-means & K-means clustering algorithms are validated using Signal to Noise Ratio (SNR) as a performance metric. Fuzzy c-means provide an enhanced detection when compared with that of k-means. The obtained SNR values prove that the FCM provides the very nearer value of the defects compared to that of K-means. Thus, fuzzy c-means with Barker coded Thermal Wave Imaging can provide effective detection of the anomaly.