上海交通大学学报(自然版) ›› 2015, Vol. 49 ›› Issue (03): 361-365.

• 金属学与金属工艺 • 上一篇    下一篇

变极性等离子弧穿孔立焊双面图像处理与特征提取

陈路路,胡明华,陈云霞,吴頔,陈华斌,林涛   

  1. (上海交通大学 材料科学与工程学院,上海 200240)
  • 收稿日期:2014-09-20 出版日期:2015-03-30 发布日期:2015-03-30
  • 基金资助:

    国家自然科学基金项目(51275301)资助

Double-Side Image Processing and Feature Extraction of  Variable Polarity Vertical-Up Plasma Arc Welding

CHEN Lulu,HU Minghua,CHEN Yunxia,WU Di,CHEN Huabin,LIN Tao   

  1. (School of Materials Science and Engineering, Shanghai Jiaotong University, Shanghai 200240, China)
  • Received:2014-09-20 Online:2015-03-30 Published:2015-03-30

摘要:

摘要:  针对变极性等离子弧穿孔立焊的双面图像进行特征信号提取,提出了一种适应性强的图像处理算法来获取焊接参数.开发了基于组合部件模型匹配和边缘窗口线性扫描的组合算法,并通过筛选法则提取特征信号,实现对焊接双面特征的有效提取.结果表明,所提出的方法能获得可靠的正面熔池及背面小孔的几何特征信息,为视觉多信号融合提供基础.

关键词: 变极性等离子弧焊, 双面图像, 小孔熔池图像, 组合部件模型

Abstract:

Abstract: In order to obtain the double-side image feature signal for variable polarity verticalup plasma arc welding, an adaptable image processing method was proposed. The tree part-based model image processing algorithm and linear scanning edges algorithm were used simultaneously, and the appropriate one was selected. The result shows that reliable geometric features of front pool and back keyhole puddle can be obtained with the proposed method, which enables it to integrate muti-visual signals.

Key words: variable polarity vertical-up plasma arc welding(VPPAW), doubleside image, keyhole puddle and welding pool, part-based model

中图分类号: