基于欧氏距离图的随机Hough变换椭圆检测方法

摘要

为了实现椭圆目标的有效检测,克服椭圆检测过程中对椭圆完整性和边缘梯度精度要求过高的缺点,提出了一种改进的随机Hough变换的椭圆检测方法.首先充分利用椭圆的轴对称特性和极点-极弦性质求取候选椭圆,有效解决了无效采样和累积问题,然后采用欧氏距离图计算椭圆边缘点的欧氏距离之和来确定真实椭圆.实验结果表明,该算法相对于RHT-3算法和CMHT算法具有检测精度高、检测速度快和抗椭圆缺失能力强的优点.%To detect the elliptical object effectively,and overcome the high requirements of edge gradient accuracy and el-lipse integrity in ellipse detection process,an improved ellipse detection method based on improved random Hough transform is proposed. The axisymmetric characteristic,and pole and polar line property are fully used to get the candidate ellipse to solve the invalid sampling and accumulation problems effectively. The Euclidean distance graph is used to calculate the sum of Eucli-dean distances of the ellipse peripheral point to determine the true ellipse. The experimental results show that,in comparison with RHT-3 algorithm and CMHT algorithm,the algorithm proposed in this paper has higher detection accuracy,faster detection speed,and stronger ability to resist the ellipse loss.