环形焊缝的双目视觉定位与跟踪研究

Study on Binocular Vision Location and Tracking of Circular Seam

为有效解决非标罐体的环形焊缝识别问题,设计了一种基于被动双目视觉的环形焊缝定位与跟踪系统。首先改善左右相机采集到图像的图像质量,运用基于霍夫变换实现动态提取感兴趣区域(Region of interest,ROI)。然后经图像处理去噪,采用基于二值矩阵去除预焊点的方法来去除预焊点对焊缝识别的干扰,运用最小二乘法补全焊缝信息,拟合焊缝直线。接着利用改进后的双目视觉模型和基于极线的立体匹配方式,将环形焊缝细分为多个焊缝小区间内的直焊缝处理,计算并得到环形焊缝上焊点的空间坐标。实验结果表明,该系统平均误差为0.14 mm,并且可以准确地识别焊缝位置,验证了系统具有良好的精度。

In order to effectively solve the circular seam identification for non-standard tanks,a circular seam location and tracking system based on passive binocular vision was designed.Firstly,the quality of the images collected by the left and right cameras is improved,and the Region of Interest(ROI)is dynamically extracted based on the Hough transform.Then image processing is used to denoise,and the method based on binary matrix removal of pre-welds is used to remove the interference of pre-welds on weld recognition.The least-square method is used to complete the weld information and fit the weld straight line.Then,with the improved binocular vision model and the stereo matching method based on the epipolar line,the circular seam is subdivided into straight welds within multiple seam sections,and the space coordinates of the weld points on the circular seam are obtained.The experimental results show that the average error of the system is 0.14 mm,and the weld position can be identified accurately,which verifies that the system has a good accuracy.