Signal processing for the detection of multiple imperfection echoes drowned in the structural noise

Area of Science:

• Materials Science
• Signal Processing
• Non-Destructive Testing

Background:

• Structural noise in materials can obscure critical defect echoes.
• Accurate detection of material imperfections is vital for safety and quality control.

Purpose of the Study:

• To develop and validate algorithms for detecting and locating multiple defect echoes within material noise.
• To assess the robustness of the proposed method against varying defect characteristics and noise levels.

Main Methods:

• Utilizing split spectrum processing combined with a multi-step approach based on Group Delay Moving Entropy (GDME).
• Investigating the method with four known defect echoes of varying positions, center frequencies, and bandwidths.
• Generating grain noise signals through both a clutter model and experimental coarse-grained materials.

Main Results:

• The GDME method demonstrates the ability to distinguish defect echoes (constant group delay) from noise (random group delay).
• Signal-to-noise ratio calculations confirm the robustness of the detection method across varied defect frequencies.
• Successful detection and localization of multiple defect echoes submerged in structural noise.

Conclusions:

• The proposed GDME-based algorithms offer a robust solution for detecting material defects in noisy environments.
• This technique enhances the reliability of non-destructive testing by improving the signal-to-noise ratio for defect identification.