Method for defect contour extraction in terahertz non-destructive testing conducted with a raster-scan THz imaging...

Area of Science:

• Non-destructive testing
• Terahertz imaging
• Signal processing

Background:

• Terahertz (THz) imaging systems, particularly frequency-modulated continuous-wave (FMCW) systems, are valuable for non-destructive testing.
• Extracting defect contours from THz FMCW interference can be challenging due to severe stripe noises.

Purpose of the Study:

• To develop a novel algorithm for accurate defect contour extraction in THz FMCW non-destructive testing.
• To validate the algorithm's effectiveness in handling noisy interference and precise defect area measurement.

Main Methods:

• Development of a neoteric algorithm utilizing a two-dimensional continuous wavelet transform.
• Application of the algorithm to results from a raster-scan frequency-modulated continuous-wave (FMCW) THz imaging system.
• Experimental validation of the developed method.

Main Results:

• The algorithm successfully extracts defect contours from THz FMCW interference corrupted by severe stripe noises.
• The relative error in defect area calculation between actual values and extracted contours is below 3.03%.
• Demonstrated effectiveness in noisy THz imaging data.

Conclusions:

• The proposed wavelet transform-based algorithm offers an effective and exact solution for defect contour extraction in THz FMCW non-destructive testing.
• The method shows high precision in quantifying defect areas, crucial for material analysis and quality control.
• This advancement improves the reliability of THz imaging for defect detection.