Current Issue Cover
基于知识的X光焊缝缺陷三维信息提取

梁德群1, 袁小川2, 杨海军3(1.大连海事大学信息工程学院,大连 116026;2.美国纽约州立大学;3.西安交通大学图象处理与识别研究所,西安 710049)

摘 要
介绍了一种焊缝X线图象缺陷三维信息提取的方法.一般在二维X线投影图上,会把沿射线方向的走向裂纹等缺陷当成气孔或夹渣,而且会忽略很薄的面积性缺陷(裂纹和未焊透),还可能把表面凹坑看成内部气孔,但从三维来考察焊缝缺陷,不仅可避免上述问题,而且可为补焊提供深度信息,为焊接工艺分析提供帮助.因而三维缺陷自动检测技术将引起焊缝无损检测工艺革命性变化和带来巨大的经济效益.由于现场条件限制着投影图的数量,为了有效地利用这些X射线投影图,文中不仅讨论了成象方法和X光源的模型,还构造了基于代数迭代法的缺陷三维信息
关键词
3-D Information Detection of Weld Seam Facts inX-Ray Image Based on Knowledge

()

Abstract
A method of 3-D information detection of weld seam facts based on X-ray image has been developed. There are problems in the fact detection on the X-ray 2-D projection of weld seam, which the cracks along the projection direction and the concave holes on the surface of weld seam may be recognized as the air holes and the thin area facts(the cracks and nonpenetrate solder) may not be detected. The problems can be avoided in the method of 3-D information detection. The paper discusses the imaging approaches and the light models. An iterative algorithm of 3-D information detection of the facts has been constructed. This algorithm has been simplified greatly by parameterization of the fact models using the form knowledge about the facts.
Keywords

订阅号|日报