基于涡流C扫描成像的缺陷量化检测技术研究

摘要

提出了基于涡流C扫描成像的缺陷检测方法,以直观准确地识别与量化缺陷参数.首先搭建涡流C扫描成像检测装置,用以测量不同尺寸和深度的裂纹试件,得到目标区域的C扫描图像;然后利用主成分分析方法提取缺陷的位置信号特征;最后在位置识别的基础上,采用Sobel算子获取缺陷轮廓,并应用污染区域与裂纹实际长度值之间的线性关系对参数估计值进行修正,提高缺陷尺寸检测精度.实验结果表明,所提出的方法可以快速识别缺陷位置、缺陷轮廓,识别精度在5% 以内.%A novel method based on eddy current C-scan imaging detection is proposed,which can extract the key information of the crack dimension.The first step is to build the eddy current C-scan imaging detection setup,and samples with different sizes and depths are detected.Then,the C-scan image of samples is obtained.Principal component analysis is adopted for feature ex-traction to detect the location of the crack in the samples.Based on the location identification,Sobel operator is utilized to extract the outline of the crack.The dimensions of the crack are corrected based on the relationship between the contaminated area and the real area of the crack to improve the prediction accuracy.The experimental results show that the method is efficient and accu-rate.