摘要
针对激光拼焊过程中的各种噪声干扰,使得焊缝图像复杂多变,本文对结构光视觉焊缝质量检测系统图像处理方法进行了深入研究。在图像预处理中,首先通过加窗处理来获得兴趣区域,采用中值滤波去除图像噪声。针对母材区域和焊缝区域对结构光反射率不同,提出了使用局部阈值来分割目标图像的方法,并使用形态学开运算进一步去除噪声干扰;在结构光条纹中心线提取过程中,使用模板法获得了条纹的边界并用几何中心法提取了条纹中心线;提出了基于焊缝灰度图像的灰度突变和拟合直线法来检测特征点的方法。试验表明,该方法具有较高的特征点检测可靠性,并且运算速度快、抗干扰能力强,具有较高实用价值。
In the seam surface quality visual inspection system of laser welding of tailored blanks,the main purpose of image processing was to extract the centerline of structured light stripe and the feature points.To the further study of image processing method of seam quality inspection system,a reasonable and efficient image processing method for seam quality inspection system was presented in this paper.In the image preprocessing,the region of interest was obtained by using adding window operation,and median filter was adopted to remove the noise interference.Local threshold segmentation was utilized to extract the structured light stripe.Also,in this processing,a template was used to detect the edges of the structured light stripe,then geometrical center method was applied to extract the centerline of structured light stripe,after that,gray scale mutation and linear fitting method based on weld grey image was presented which was used to detect the feature points.The results showed that the proposed method had the advantages of higher reliability and stronger anti-interference ability,laid a foundation for the realization of on-line inspection of seam surface quality visual inspection system.
出处
《焊接技术》
北大核心
2010年第4期56-61,共6页
Welding Technology
基金
国家自然科学基金项目(50705095)
关键词
激光拼焊
焊缝图像
结构光条纹
特征点检测
laser welding of tailored blanks,seam image,structured light stripe,feature points detecting
