摘要
为了采用图像处理与模式识别的方法对铝合金的焊接质量进行智能控制 ,必须获取铝合金焊接熔池的清晰图像 ,然后提取焊接熔池的信息 .在对铝合金焊接光谱分析的基础上 ,以焊接电弧光中的连续光谱作为光源 ,采用宽带滤波方法设计的全新的光学图像传感器 ,利用普通工业CCD摄像机获得了画面清晰、质量稳定的铝合金焊接熔池区域图像 .针对所获得的铝合金焊接熔池图像中干扰分配的不均衡性 ,采用新型分区滤波图像预处理算法 ,得到了铝合金焊接熔池的理想特征参数 .该算法具有抗干扰强、速度快的特点 。
In order to control the quality of welding by image processing and pattern recognition, the clear image of the welding pool of aluminum alloy and the information of the welding pool must be obtained. A completely new optical sensor is designed by the analysis to the light spectrum of welding of aluminum alloy which is on the broad band filtering and the welding arc light is used as resource during the work. It greatly reduces the noise from the discontinuous spectrum in the arc and adapts to the seriously fluctuating welding current during the process and obtains the clear images of the welding pool using common industrial CCD camera. In view of these things that noise distribution is not even in the aluminum alloy's welding pool image, a pretreatment arithmetic named subarea filter is applied in the images proceeding. With the new algorithm, the desirable characteristic parameters of the welding pool of aluminum alloy's AC tungsten inert gases welding are obtained. The arithmetic's anti interference ability is high and operation speed is quick.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2002年第1期5-8,共4页
Journal of Shanghai Jiaotong University
基金
中国航天科技总公司重大技术改造项目 ( 0 5 0 12 0 0 0 3)
关键词
铝合金
交流钨极氩弧焊
熔池
图像传感
aluminum alloy
AC tungsten inert gases welding
spectral analysis
image sensing
