A comparative study of principal component analysis and independent component analysis in eddy current pulsed thermography data processing.
A comparative study of principal component analysis and independent component analysis in eddy current pulsed thermography data processing.
复制标题
DOI：
10.1063/1.4823521
复制
发表时间：
2013-10
期刊：
影响因子：
--
通讯作者：
L. Bai;B. Gao;S. Tian;Yuhua Cheng;Yifan Chen;G. Tian;W. Woo
中科院分区：
文献类型：
--
作者：
L. Bai;B. Gao;S. Tian;Yuhua Cheng;Yifan Chen;G. Tian;W. Woo
涡流脉冲热像仪(ECPT)是一种新兴的无损检测与评价技术，已被广泛应用于各种材料中。横向热扩散导致缺陷区域与无缺陷区域的温差减小。为了增强缺陷的对比度，近年来人们提出了不同的统计方法，如主成分分析和独立成分分析，用于热像图像序列的处理。然而，这两种算法的实现都缺乏直接和详细的独立比较。本文的目的是比较这两种方法，并确定ECPT数据中缺陷对比度增强的最佳技术。进行了人工裂纹和热疲劳天然裂纹检测的验证实验。
Eddy Current Pulsed Thermography (ECPT), an emerging Non-Destructive Testing and Evaluation technique, has been applied for a wide range of materials. The lateral heat diffusion leads to decreasing of temperature contrast between defect and defect-free area. To enhance the flaw contrast, different statistical methods, such as Principal Component Analysis and Independent Component Analysis, have been proposed for thermography image sequences processing in recent years. However, there is lack of direct and detailed independent comparisons in both algorithm implementations. The aim of this article is to compare the two methods and to determine the optimized technique for flaw contrast enhancement in ECPT data. Verification experiments are conducted on artificial and thermal fatigue nature crack detection.