文档介绍：摘要
摘要
接质量和生产效率,焊缝跟踪是实现焊接自动化的必要技术。本文设计了
基于DSP的焊缝跟踪系统,D进行焊缝图像的采集,利用高速DSP进行
图像处理、偏差量计算以及焊缝纠偏过程控制。
鉴于大量的飞溅和弧光干扰了清晰图像的采集,本文采用相应的滤光系统
将波长在650 ±10 nm以外的光滤除,余留的光作为光源;D
和激光器的几何位置。经过滤波系统后视觉传感器可采集焊前焊缝区的图像,
获得信息丰富直观的焊缝信息。
通过对焊缝原始图像采取中值滤波预处理技术,得到了无大噪声干扰的焊
缝区图像。对焊缝图像进行模糊边缘检测,可以得到焊枪前方十分清晰的激光
线边缘。通过激光线边缘,可以提取出特征线,得到焊缝位置参数并计算出焊
缝偏差量。本文将模糊边缘检测应用到焊缝图像处理中,方法简单,边界图像
清晰完整;运用高速的DSP系统进行图像的处理,保证了处理的实时性。
在确定的图像处理基础上,本文采用了比例控制和模糊控制技术相
结合的 Fuzzy-P 控制方法。当偏差较大时,采用比例控制提高系统的响
应速度,加快响应过程;当偏差较小时应用模糊控制,获得稳定的系统
误差。本文把两种控制方法结合起来构成了兼有两者优点的分段模糊比
例控制器。
为了验证系统的工作性能,本文对不同焊缝形状和不同偏差量的情况进行
了纠偏试验,试验表明跟踪误差小于± 。进一步证实了采用机器视觉、
图像处理方法以及智能控制可以满足焊缝自动跟踪的要求。
关键词:焊缝跟踪;DSP;模糊边缘检测;图像处理;Fuzzy-P 控制.
I
Abstract
Abstract
Welding plays an important role in material processing. Welding automation can
improve welding quality and productivity greatly. Seam tracking is an essential
technique to fulfill welding automation.
In this paper, welding seam tracking system based on DS P is designed D
is Applied to gather welding seam image and DSP to image processing、account warp
& intelligent control.
It is very difficult to detect clear and steady image owing to the noise
interferences that exist in the nearby weld line caused by vast arc, spatter, etc. This
paper corresponding filter system is applied to remove the light which wavelength out
the range of 650 ±10 nm; and the remains is kept as light source. After affirms the
relative geometrical position D and laser by experiments, it can shot weld line by
visual sensor, which contains abundant and objective information.
Through adopting the median filter for originality seam image, get the seam area
image which has no big noise. During fuzzy edge detection, clear laser edge can be
gotten. According to laser edge, the characteristic line and the parameters of the seam
position and the seam accurate warp information can be obtained. In this thesis, fuzzy
logic is applied to image detects and it has the ch