论文部分内容阅读
针对射线探伤中采集的焊缝图像存在对比度低,噪声干扰及边界模糊等特点,提出了一种基于多结构元素形态学的射线焊接图像缺陷边缘检测方法。在对原始射线焊接图像进行预处理、焊道提取后,选取了多尺度多结构元素,运用灰度形态学方法对焊接图像进行缺陷边缘检测与提取。试验结果表明,对比传统的边缘检测算法,该方法能有效提取图像中的焊接缺陷边缘,且其连续性和完整性较好。
In view of the low contrast, noisy interference and fuzzy boundaries of the weld images collected during the radiofrequency detection, an edge detection method based on morphologies of multi-structure elements is proposed. After preprocessing the original welding image and extracting the weld bead, multi-scale and multi-structure elements are selected, and gray morphological method is used to detect and extract the defect edge of the welding image. The experimental results show that compared with the traditional edge detection algorithm, this method can effectively extract the edge of the welding defects, and the continuity and integrity of the method is good.