减影提取焊缝缺陷算法  

Weld Defect Extraction Algorithms Based on Subtraction Angiography Method

作　　者：王鹏[1] 王婧[1] 杜卫东 

机构地区：[1]西安工业大学电子信息工程学院,西安710021 [2]山东省特种设备检验研究院,济南250101

出　　处：《西安工业大学学报》2013年第10期856-860,共5页Journal of Xi’an Technological University

基　　金：陕西省科技厅工业攻关项目(2013K07-15);陕西省教育厅专项科研计划资助项目(11JK0921);山东省科技厅科技发展计划项目(2010GWZ20402);山东省质检总局科研项目(2009KYZ22);质检工艺性行业科研专项(201010030)

摘　　要：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的问题,提出了一种基于减影法的焊缝缺陷智能检测提取方法.首先使用模糊增强方法对焊缝图像进行灰度变换,而后利用像素标记提取焊缝并保留其原始灰度值,再对提取到的焊缝滤波,去除其内部缺陷,衍生出背景图像,最后根据减影法得到缺陷,并且计算出缺陷的特征参数.结果表明,经上述方法检测后的缺陷图像,对比待测焊缝原图,可以较为准确地检测出焊缝中存在的缺陷,具有较好地自适应性和实用性.For the difficulty to accurately extract weld defect by the X-ray automatic detection technology, an intelligent detection and extraction method for weld defects based on the subtraction angiography is put forward. First fuzzy enhancement method is used to weld image gray-scale transform, then pixel markers is used to extract weld and the original grey value is retained. Tthe extraction of the weld seam is filtered to remove the internal defects and to create the background image. By the subtraction angiography method, the defects are obtained and the characteristic parameters of defects are calculated. The results show that compared with the original image of the weld seam,the defect image is processed by the above method accurately. It has good adaptability and practicability.

关 键 词：焊缝缺陷 模糊理论 标记 减影法 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]