AI enhanced strong-field terahertz spectral detection and imaging

人工智能增强的强场太赫兹光谱探测与成像

阅读:1

Abstract

Terahertz (THz) waves demonstrate advantages including broad absolute bandwidth, low photon energy, and superior penetration capability, showing significant potential in nondestructive testing and spectroscopic imaging applications. However, current techniques face challenges including unknown material refractive indices, pulse signal aliasing, and time-consuming detection processes, which hinder the advancement and commercialization of THz-based nondestructive testing. This work presents a nondestructive testing system employing an enhanced 4-inch spintronic strong-field THz emitter, integrated with neural-network-assisted thickness prediction and contour detection. The system achieves micrometer-scale accuracy for ultrathin materials and rapid defect identification in large-scale samples. Experimental results demonstrate 99.8% measurement accuracy within ±8 μm for micrometer-scale samples, with submillimeter-level depth resolution for defect characterization and contour imaging. This technology demonstrates scalability and high flexibility, laying the foundation for thickness measurement of multilayer composite coatings and internal defect detection in complex materials, showing broad prospects in surface coating processing, component maintenance, and cultural heritage preservation.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。