一种基于结构光的V型焊缝实时图像处理方法

The real-time image processing V-shaped seam method based on a structure light

在焊缝跟踪系统中,需要实时准确获得焊缝中心线的位置,就必须设计一种基于结构光的V型焊缝中心线的提取方法。对焊缝进行中值滤波去除图像噪声,采用自适应阈值分割的方法对图像进行二值分割,采用二值形态学的边缘检测方法对二值后的图像进行孤点滤波和边缘检测,采用对边缘上下边界点求取平均值方法提取结构光的中心线,采用基于斜率分析和最小二乘法相结合的检测的方法对图像进行特征检测获得图像的特征信息。通过实践表明,所采用的方法能够正确地提取焊缝特征信息,具有很强的抗干扰性能,能够满足焊缝跟踪系统的实时性要求。

In the seam tracking system,need real-time access to accurate weld centerline position,we must design an extraction V-weld centerline method based on structured light.Weld median filter to remove image noise,using adaptive threshold segmentation method for image binary segmentation,edge detection method using binary image binary morphology were solitary point after filtering and edge detection,and then using the centerline of the upper and lower edges of the boundary point method of obtaining the average light extraction structures,slope analysis method based on the least squares method and the detection of the image feature detection feature information obtained images.Through practice shows that the method used to correctly extract feature information weld,has a strong anti-jamming capability to meet the requirements of real-time seam tracking system.