基于标准控制场的机器视觉测量方法

Vision Measurement Based on Standard Control Field

针对常规机器视觉测量中累积误差过大的问题,提出了一种基于外加标准控制场的拼接模型,并对控制场中控制点的分布方式、公共点个数对测量精度的影响以及控制场的加工可行性等进行分析,找到了最合适的控制场设计形式。实验结果表明,基于外加控制场的拼接模型在汽车大梁上装配孔孔位测量中适应性很好,消除了累积误差的影响,使整体测量精度大大提高,同时还能弥补因工件自身几何特征稀少而不能进行图像拼接的不足。

In order to solve the problem of excessive accumulation of errors in conventional machine vision measurement, a new image mosaic method based on control field is presented. Firstly, the distribution and number of control points in control field which impact on measurement accuracy are analyzed. Then processing feasibility of the control field is discussed. Finally, the most suitable design form of the control field is obtained. Experimental results show that this new mosaic method has good adaptability in the measurement of mounting holes on car beam. It not only eliminates the effect of cumulative error, and the overall measurement accuracy is greatly improved, but also compensates for the lack of image stitching due to the scarce characteristics on the workpiece itself.