基于Radon变换的高精度Mark点圆定位算法的改进

The Improvement Algorithm of High Precision Mark Points Detecting Circle Location Based on Radon Transform

Mark点的圆定位技术在印刷、电子制造、激光加工等行业有着广泛的应用。为提高定位精度而一味地提高Mark点的印刷质量会带来成本的上升,通过算法改进消除Mark点边缘瑕疵和圆缺省的影响,可以提高定位精度。在提取Mark点圆图像的边缘后,采用Radon变换将定位圆f(x),y变换到f(ρ),θ,通过抽取每一θ角上的ρ的最值,并与标准圆对应的最值比较,按一个设定的阀值判定是否噪声,从而区分定位圆的真实边缘点和噪声,再通过最小二剩法对真实边缘点拟合求出圆心。实验证明,本改进方法针对有缺陷的定位圆有较高的定位精度与计算稳定性。

The Mark points detecting circle location technology is widely used in the industry application of printing, the electronicmanufacturing,the Laser machining and so on. Manufacturing cost would be increased if only rising the printing quality of the mark pointscircle. The precision of detecting circle location can be improved by using the algorithm which eliminating effects on mark points circleedge defectives and circle lacks. After the mark points circle edge extracting,using the Radon transform employed to transmute the circlef（ x）,y tof（ ρ）,θ,finding the peak point at the every and compare this point value with the corresponding value of the standard circle,judge the point is or not noise by the value difference large then a threshold value or not. Then use minimal two remaining methodfit the real circle edge points. The experiments prove that the improvement vision circle location algorithm has more location precision andstability.