基于抛物线拟合的TOFD图像缺陷检测

针对TOFD(time of flight diffraction)检测的D扫描图像,提出了一种新的图像缺陷检测方法.采用基于能量分布的杂波抑制算法去除直通波及底面反射波,通过二值化处理、边缘检测、腐蚀以及去除小聚类等预处理方法去除背景杂波,然后,提出改进的最小二乘法提取缺陷的两端分段拟合抛物线,并对特征点及其拟合点的纵坐标向量进行相关性分析,最后根据缺陷特征点与抛物线拟合的相关系数检测图像缺陷.结果表明,该方法准确有效.

Defect detection of TOFD D scanning image based on parabola Htting

A new method was introduced to accurately detect the defects of time of flight diffraction (TOFD) D scanning image. First, the clutter suppression algorithm based on energy distribution was used to remove direct wave and reflected bottom wave. The background clutter could be removed through algorithms of binarization processing, edge detection,corrosion, clustering and removal of small clusters. Then, a parabola was fitted through the least square method, and ordinate vectors of feature points and fitted points were on the correlation processing. With the matching degree of defects and parabola, the defects could be detected. The results demonstrate that the proposed method was effective and accurate.