基于模糊理论的焊缝缺陷图像识别算法研究

为提高无损检测中焊缝缺陷射线检测效率、降低人工操作强度,针对射线胶片焊缝缺陷自动检测技术普遍存在难以自动准确识别缺陷问题,充分利用焊缝图像中各种信息,提出了一种焊缝缺陷智能检测识别方法。基于自动选择阈值,使用模糊集合进行灰度变换以增强图像对比度;利用最大类间方差算法(OTSU)求解自适应阈值,检测并提取焊缝边缘;依据缺陷几何特征识别不同缺陷。结果表明,该系统可以较为准确地检测各种焊缝图像的边界和缺陷,具有较好地适应性和实用性。

Based on the theory of fuzzy weld defect image recognition algorithms

In order to improve the detection efficiency of the nondestructive testing of weld defect ray, reduce labor intensity of operation, according to the common problems of the ray automatic detection technology for weld defect is difficult to accurately extract, make full use of the information of weld images, put forward a intelligent automatic detection method for weld defects. Based on the automatic threshold, use fuzzy set for gray transform to enhance image contrast; use the OTSU to solve adaptive threshold value to identify and extract the weld edge; according to the geometry characteristics identification different defects. The results show that the method can accurately test the boundaries and defects of all kinds of weld images, has a good adaptability and practicability.