激光视觉焊缝跟踪中图像二值化处理

Image Binarization Processing in Laser Vision Seam Tracking

焊缝图像的二值化处理是实现焊缝视觉跟踪的一个重要环节.针对激光视觉传感焊缝图像,通过最小二乘法建立了Otsu法所选阈值与焊缝激光带区域灰度平均值的直线回归方程,从而得到了焊缝图像二值化处理的最佳阈值.利用该方法对焊缝图像进行二值化处理,并与传统二值化处理方法的结果进行了对比.结果表明,该二值化阈值选取方法可以有效地选取出适合焊缝图像二值化处理的阈值,处理结果良好,更利于后续的焊缝特征点提取.

The binarization processing of seam image is an important step to realizing the vision seam tracking.Aim-ing at the seam image of laser vision sensing,the linear regression equation of threshold obtained using Otsu method and average gray value of laser region was concluded based on least square method,hence the optimal threshold of seam image binarization was gained.The binarization processing result of the calculated threshold was compared with that of conventional binarization approach.The results show that the suitable threshold of seam image binarization processing can be automatically obtained,and the seam image binarization result is beneficial to the subsequent im-age feature point extraction.