高级检索

基于超声C扫图像反卷积的棒材缺陷定量方法

Flaw sizing method for rod workpieces based on deconvolution of ultrasonic C-scan image

  • 摘要: 缺陷定量评价一直是无损检测领域研究的重点,超声波的扩散会导致C扫描图像边缘模糊,影响缺陷定量的准确性。为了提高棒材缺陷定量的准确性,提出了一种棒材超声C扫描图像反卷积方法。基于多元高斯声束模型(multi-gaussian beam model, MGB),根据超声波在曲面界面的传播规律,推导了棒材水浸超声成像系统的点扩散函数(point spread function, PSF)。利用四轴超声水浸检测系统对含有平底孔的尼龙棒进行了检测,得到了C扫描图像,然后用Richardson-Lucy(RL)迭代算法对C扫描图像进行反卷积处理。使用6 dB下降法分别对原始图像和反卷积处理后的图像进行定量分析。结果表明,反卷积处理后定量误差更小,特别是将直径1.5 mm平底孔的误差由88%降低到了18.7%。

     

    Abstract: Flaw sizing is the focus of research in the flied of non-destructive testing. The diffusion of ultrasonic waves will cause edge blur of C-scan images, which affects the accuracy of flaw sizing. Based on the multi-Gaussian beam model, according to the propagation law of the ultrasonic wave at a curved interface, the point spread function (PSF) of the rod water-immersion ultrasonic imaging system is deduced. The nylon rod with flat-bottomed holes is scanned by a four-axis ultrasonic water-immersion testing system to obtain the C-scan image, and then the C-scan image is deconvoluted with the Richardson-Lucy iterative algorithm. The 6 dB drop technique is used to quantitatively analyze the original image and the deconvoluted image. The results indicate that the proposed deconvolution method exhibits a higher sizing accuracy. In particular, the error of the 1.5 mm diameter flat-bottom hole is reduced from 88% to 18.7%.

     

/

返回文章
返回