基于希尔伯特变换和功率谱估计的薄缺陷太赫兹检测
投稿时间:2021-04-17  修订日期:2021-06-15  点此下载全文
引用本文:
摘要点击次数: 171
全文下载次数: 0
作者单位邮编
张瑾 吉林大学仪器科学与电气工程学院 130061
谭冰冲 吉林大学仪器科学与电气工程学院 
陶星竹 吉林大学仪器科学与电气工程学院 
徐成城 吉林大学仪器科学与电气工程学院 
常天英 吉林大学仪器科学与电气工程学院 
崔洪亮 吉林大学仪器科学与电气工程学院 
王洁 吉林大学仪器科学与电气工程学院 130061
基金项目:山东省自然科学基金项目 (ZR2020KF007)
中文摘要:提出将希尔伯特变换和功率谱估计相结合的光谱分析算法,对太赫兹反射时域波形进行处理,并将该算法应用于太赫兹时域光谱成像,将缺陷厚度和图像灰度关联,实现同时对玻璃纤维层压板内部缺陷厚度、位置和形状的成像检测。实验结果表明:将多重信号分类谱估计、自回归谱估计和希尔伯特变换结合时,能成功区分厚度为0.08 mm缺陷上下表面反射脉冲,反射脉冲的时间分辨率高于0.5 ps,缺陷厚度的检测误差低于0.03 mm。
中文关键词:太赫兹时域光谱  厚度检测  希尔伯特变换  功率谱估计  玻璃纤维层压板
 
Terahertz Detection of Thin Defects based on Hilbert Transform and Power Spectrum Estimation
Abstract:A spectral analysis algorithm based on the combination of Hilbert transform and power spectrum estimation has been proposed, and the terahertz reflection time domain waveform was processed. At the same time, the algorithm was applied to terahertz time domain spectroscopy imaging, defect thickness was correlated with image gray level, and the thickness, position and shape of defects in glass fiber laminate can be detected by imaging simultaneously. The experimental results show that when the multi-signal classification spectrum estimation and autoregressive spectrum estimation are combined with Hilbert transform, the reflected pulses between upper and lower surfaces of defect with thickness of 0.08 mm can be successfully distinguished, the time resolution of reflected pulses is higher than 0.5 ps, and the detection error of defect thickness is less than 0.03 mm.
keywords:Terahertz time domain spectroscopy,Thickness estimation,Hilbert transform,Power spectrum estimation, Glass fiber laminate
  HTML  查看/发表评论  下载PDF阅读器

《红外与毫米波学报》编辑部