基于结构光视觉的白车身覆盖件间隙面差测量方法

摘要

Aiming at problem that manual measurement of gap and flush of body-in-white is imprecision and inefficiency,adopt structured light vision measuremnent method to replace manual measurement.In the recognition of feature points for the gap gauge,calculate absolute differential for each point on the centerline of the structured light stripe,then,search for differential extremal points and treat them as feature points for gap gauge.For the recognition of feature points of flush gauge,use feature points localization algorithm based on extremal curvature of discrete curve.Experimental results indicate that the proposed algorithm has a higher precision for calculating the gap and flush of body-in-white,and can satisfy the requirement of precision for car body factories.%针对白车身表面覆盖件间隙面差人工测量精度差、效率低等问题,提出了利用结构光视觉测量方法代替人工测量.在间隙特征点识别中,计算光纹中心线上各点的绝对差分值,搜索差分极值点并将该极值点作为特征点;在面差特征点识别中,采用基于离散曲线曲率极值点的特征点定位算法.实验结果表明:间隙面差测量方法测量精度较高,能够满足整车厂对白车身间隙面差测量精度的要求.