Real-time imaging for arc welding
暂无分享,去创建一个
Abstract ESPRIT (CIM) Project 9 has taken automated welding systems as instances of 'machines' for which have been developed local, intelligent, corrective systems. Two widely different systems have been studied: single-pass welding, representative of the thin-sheet/high-volume industry (e.g. automotive), and multi-pass welding, representative of the heavyweight/low-volume industry (e.g. power generation). Vision sensory systems have been used to effect control of the welding systems. The problems of signal co-ordinating and analysis to give meaningful, real-time, output in the absence or presence of an arc, have been addressed. The single-pass system provides user programmable process control, whilst the multi-pass system derives its process control from algorithms embedded in the system: both systems provide full positional control. This paper gives an overview of work undertaken over a 5-year period (1983/88) under the ESPRIT (CIM) Programme (Project Number 9), which has been fully documented in the bi...
[1] Dana H. Ballard,et al. Computer Vision , 1982 .
[2] Azriel Rosenfeld,et al. Digital Picture Processing , 1976 .
[3] George G. Dodd,et al. Computer Vision and Sensor-Based Robots , 1979 .
[4] Katsunori Inoue,et al. Automatic Control of Horizontal Narrow Gap Welding (Report II) : Detection and Measurement by Image Processing(Welding Physics, Processes & Instruments) , 1980 .