Automated Internal Defect Identification and Localization Based on a Near-Field SAR Millimeter-Wave Imaging System
Fast and cost-effective detection of internal defects is essential for structural integrity inspection in various applications such as manufacturing, construction, and aerospace. Current internal non-destructive testing (NDT) methods, such as computed tomography, can be costly, time-consuming, and c...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10872942/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Fast and cost-effective detection of internal defects is essential for structural integrity inspection in various applications such as manufacturing, construction, and aerospace. Current internal non-destructive testing (NDT) methods, such as computed tomography, can be costly, time-consuming, and constrained by object size. The lightweight and affordable millimeter-wave (mmWave) radar has demonstrated the capability of detecting objects beneath surfaces or obstacles through generated near-field synthetic aperture radar (SAR) images. However, its use in precise internal inspection has not been fully explored due to the significant noise in the SAR images and high rate of false identification. To enable accurate and fast inspections using the mmWave radar, this work establishes a robust and automated internal defect detection system. It employs a compact mmWave radar system mounted on a stable rail-based scanning mechanism, generating high-resolution near-field SAR images with an enhanced signal-to-noise ratio through denoising. For fast and accurate detection, an automated defect localization algorithm is developed. The accuracy of detection is ensured by modeling and separating internal defects from disturbances introduced by the scanning mechanism and noise. Experiments were conducted using 3D-printed blocks with synthetic defects to demonstrate the detection capability of the proposed system. Internal defects were accurately detected across variations in shape, material permittivity, depth, and size. The proposed method achieved an average accuracy of 91.7%, outperforming existing methods. The compact design of the radar system enables seamless integration with larger scanning systems, while the automated detection algorithm can be readily implemented as a software module within existing sensing systems. This integration of hardware and software components yields a versatile, low-cost framework for rapid internal health inspection that adapts to various industrial applications and object sizes. |
|---|---|
| ISSN: | 2169-3536 |