Improving the nondestructive thermography inspection results for detection of circular defects in coated metals using principal component analysis

Coating metals as a technique to increase the corrosion resistance of industrial parts is very practical and common in various industries, especially oil and gas, petrochemical and marine industries. Corrosion characterization in coated metals is a major challenge in the field of Non-destructive testing (NDT). In this work, 20 flat bottom holes (FBH) with diameters of 10, 8, 6, 4, 2 mm were drilled to distances from the test piece surface ranging between 0.5 and 2 mm on the St37 steel plate with epoxy coating simulate corrosion defect. Step heating thermography (SHT) has been applied to evaluate corrosion defects. Median filter and Gaussian low-pass filter are used for noise reduction and smoothing the thermal images. Principal component analysis (PCA) has been applied to raw thermal images to process thermographic signals and improve the defect detection capability. Results showed that after applying signal processing methods, the number of the detected defects increased from 12 to 14 holes. Also the signal-to-noise ratio (SNR) has more than quadrupled.