• CN:11-2187/TH
  • ISSN:0577-6686

›› 1996, Vol. 32 ›› Issue (6): 31-36,5.

• 论文 • 上一篇    下一篇

扫码分享

基于焊缝CCD图象模式特征的焊缝轨迹识别

张华;陈强;金建敏;潘际銮   

  1. 清华大学
  • 发布日期:1996-11-01

WELDING SEAM TRACK RECOGNITION BASED ON WELDING SEAM CCD IMAGE PATTERN FEATURE

Zhang Hua;Chen Qiang;Jin Jianmin;Pan Jiluan   

  1. Qinghua University
  • Published:1996-11-01

摘要: 提出了一种基于焊缝CCD图象模式特征的焊缝轨迹识别的新算法,其基本思想是将焊缝等分成为段,每段的灰度分布用特征向量来描述,利用前一段的特征向量来识别下一段的特征向量,从而实现焊缝的识别。这种算法具有对各种坡口快速、准确地识别并能自适应局部和整体噪声等特点,可用于机器人无辅助光源焊接焊缝轨迹预规划,并对实时焊缝轨迹进行了识别。

关键词: 焊缝跟踪, 机器人弧焊, 模式识别, 图象处理, 预规划

Abstract: A new algorithm of welding seam track recognition based on welding seam CCD image pattern feature is presented for automatic welding seam tracking. The principle of the new algorithm is to divide welding seam into many sections, each of which is descripted with feature vector, to recognize next feature vector by former feature vector and to recognize the welding seam. The new method has advantage of fast speed, exact tracking and adaptive local and whole noise. It can be used on the trajectory preplan of robotic are welding seam and seam real-time tracking.

Key words: Image processing, Pattern recognization, Preplan, Robot arc welding, Welding seam tracking