摘要
提出一种基于信息融合优化搜索技术的X射线图像焊缝缺陷检测方法。在对X射线焊缝图像进行预处理的基础上,使用垂直焊缝方向投影的波形分析法进行焊缝区域分割,采用图像信息融合技术,综合不同判据获得启发式状态空间优化搜索的启发信息,对搜索过程进行优化,从而快速准确地检测出X射线图像中的缺陷区域。试验结果表明,提出的算法能够快速有效地检测出焊缝X射线图像中的缺陷,取得较好效果。
A new method of defect detection X-ray images of weld using optimized heuristic search based on image information fusion was introduced.In order to achieve real-time defects detection,weld segmentation using wave analysis was performed just after the pre-processing of the image,then image information fusion technology was used to set up searching base with revelatory knowledge,finally heuristic search was used to locate all the pixels in defects and the pixels were grouped by adjacency characteristic.Each group of pixels corresponds to certain defect in the weld.Experiments show that the mentioned algorithm can detect defects in X-ray images of the weld with great efficiency and reliability.The algorithm can be used for industrial applications of real-time weld defects detection in X-ray non destructive testing.
出处
《焊接学报》
EI
CAS
CSCD
北大核心
2007年第2期29-32,37,共5页
Transactions of The China Welding Institution
基金
高等学校博士学科点专项科研基金资助项目(20050003041)
关键词
缺陷检测
X射线
图像处理
信息融合
优化搜索
defects detection
X-ray
image processing
information fusion
optimized search
weld
