上海交通大学学报

上海交通大学学报（自然版） ›› 2015, Vol. 49 ›› Issue (01): 44-48.

• 金属学与金属工艺 • 上一篇    下一篇

一种基于立体视觉的焊缝成形质量评价方法

张鹏贤1，李明1，冯毅2   

  1. （1.兰州理工大学 省部共建有色金属先进加工与再利用国家重点实验室， 兰州 730050；2.兰州工业学院 材料工程系， 兰州 730050）
  • 收稿日期:2014-03-05
  • 基金资助:

    国家自然科学基金资助项目（50775105）

An Evaluation Method of Weld Formation Based on Binocular Stereo Vision

ZHANG Pengxian1,LI Ming1,FENG Yi2   

  1. (1. State Key Laboratory of Advanced Processing and Recycling of NonFerrous Metals, Lanzhou University of Technology, Lanzhou 730050, China;2. School of Materials Engineering, Lanzhou Institute of Technology, Lanzhou 730050, China)
  • Received:2014-03-05

摘要：

摘要：  提出了一种基于立体视觉检测而实现焊缝成形质量评价的新方法.从采集的焊缝原始图像中分割出表征成形特征的焊缝图像，在确定相似性测度函数、匹配窗口大小和相似性阈值的基础上，建立了基于图像灰度的焊缝区域匹配算法；由视差图建立了焊缝轮廓的三维坐标，完成了焊宽、余高的提取和焊缝的三维重构；依据焊缝外观尺寸的检验标准(GB10854—89)建立了焊宽最大差、余高最大差、填充度和直线度等焊缝成形质量参数的评价指标.实际验证结果表明，所建立的焊缝视觉评价方法和指标可以快捷有效地评判焊缝成形的质量.

关键词: 焊缝成形, 双目立体视觉, 成形参数, 评价指标

Abstract:

Abstract： A new method was proposed to evaluate weld formation based on binocular stereo vision. First, visual images that contain all information of weld formation were seperated from the images acquiced. With similarity measurement function, and with the size of matching window and similarity threshold being confirmed, an area matching algorithms was established based on gray value to measure weld formation. Next, three dimensional coordinates of weld formation was established by means of matching disparity map, based on which the extraction of weld width and weld hight and the reconstruction of weld formation images were accomplished in three dimension. Finally, according to the requirement of Chinese Standard GB10854—89, for the size examination of steel structure, the indexes to evaluate weld external quality were set up. The four indexes are the difference between maximum and minimum of weld width, the difference between maximum and minimum of weld height, the full degree of weld filled and straightness of weld formation. Weld test results show that weld formation can be evaluated by means of visual measurement method and the indexes.
Key words：

Key words:  weld formation, binocular stereo vision (BSV), forming parameters, evaluation indexs

中图分类号: