摘要
分析了传统Hough变换、随机Hough变换在检测圆及缺损圆时存在的问题,在总结了经典的Hough算法和其他研究人员提出的其它改进算法后,利用圆的几何特征,先检测出缺损圆的一个圆心坐标参数,然后根据推导出的半径检测公式对图像中的点到与圆心具有一个相同坐标参数的边缘点(即过圆心与X轴或者Y轴平行的直线与圆的交点)的距离与该点和边缘点的水平或者垂直坐标差值的关系统计半径大小,进而确定该缺损圆的所有参数。实验及分析验证了该方法降低了Hough变换的维度,比传统的Hough变换在检测缺损圆的边缘以及定位的问题上性能有较大的提高。
After analyzing the problems in the traditional methods and randomized Hough transforms to detect circle and damage circle, and summarizing the classical Hough trans- forms and other improved algorithms proposed by many researchers, an algorithm is pro- posed, which uses circle' s basic geometric features that the center of two symmetrical points in circle is the circle center of the circle, to detect one parameter of the circle center firstly , and then the radius of the defect circle is deduced. Simulation results show that all parameter of the defect circle is determined, which decreases the dimension of the trans- form.
出处
《沈阳理工大学学报》
CAS
2013年第3期52-56,65,共6页
Journal of Shenyang Ligong University
基金
辽宁省教育厅计划项目(2009A618)