X射线焊缝缺陷自动识别系统的设计与开发  

Design and development of automatic X-ray identification expert system of weld defects

作　　者：郭俊磊[1] 魏艳红[1] 占小红[1] 梁硼[1] 

机构地区：[1]南京航空航天大学,211100

出　　处：《焊接》2013年第5期36-39,69-70,共4页Welding & Joining

摘　　要：为了增加焊接检测技术的客观性、规范性,提高焊接检测效率,采用计算机技术对焊缝进行缺陷检测是有效提高方法之一。由于x射线焊缝检测的底片数字化后存在信噪比低、背景起伏大等问题,使得焊接缺陷的自动提取与识别存在较大困难。针对上述问题系统地阐述了x射线焊缝缺陷自动识别系统的开发方案,并在Windows操作系统上采用C++语言实现系统的开发。同时,利用该系统成功地将焊接缺陷从焊缝中提取出来。试验结果表明,文中所提出的x射线焊缝缺陷自动识别系统的设计方案是可行的,系统能够快速准确地识别出焊缝并提取出缺陷,具有很高的实际应用价值。It is the most efficient way to enhance the identification of weld defects based on the computer technology,which can help to improve the objectivity and normative standard in weld defects detection.The automatic recognition and identification of weld defects is quite difficult for the existing problems of low signal- to- noise ratio and high background fluctuation in digitized X- ray film.Concerning of the current problems,an expert system for weld defects detection is designed and developed with MS Visual C ++ with the function of automatic identification of weld defects.The feasibility and application value of the system were confirmed by the test result that the weld defects can be detected in great speed and accuracy.

关 键 词：图像处理 焊接缺陷 缺陷提取与识别 模糊神经元网络 专家系统 

分 类 号：TG115.28[金属学及工艺—物理冶金]