Real-time 3D imaging with Fourier-domain algorithms and matrix arrays applied to non-destructive testing

Area of Science:

• Materials Science
• Non-Destructive Testing
• Ultrasonic Imaging

Background:

• Real-time 3D ultrasound imaging is crucial for Non-Destructive Testing (NDT).
• Traditional delay-and-sum algorithms for Total Focusing Method (TFM) and Plane Wave Imaging (PWI) are computationally intensive, limiting frame rates.
• Fourier-domain algorithms offer higher frame rates but can require significant storage.

Purpose of the Study:

• To implement and compare real-time 3D imaging methods: TFM, PWI, and their Fourier-domain counterparts (k-TFM, k-PWI).
• To evaluate image quality, frame rates, and memory requirements of these methods.
• To propose and assess a novel implementation of k-TFM and k-PWI to reduce memory footprint.

Main Methods:

• Implementation of time-domain TFM and PWI algorithms.
• Implementation of Fourier-domain k-TFM and k-PWI algorithms.
• Development of a memory-efficient implementation for k-TFM and k-PWI.
• Real-time evaluation using a stainless steel block with additive manufactured porosities.

Main Results:

• Fourier-domain algorithms (k-TFM, k-PWI) achieved higher frame rates compared to time-domain TFM and PWI.
• The new implementation significantly reduced memory requirements for k-TFM and k-PWI.
• All four methods successfully imaged the 3D porosity network, with variations in image quality and speed.

Conclusions:

• Fourier-domain algorithms provide a viable path to real-time 3D ultrasound imaging in NDT.
• The proposed memory reduction technique enhances the practicality of k-TFM and k-PWI.
• This research advances NDT capabilities for characterizing complex internal structures.