摘要
为实现快速、高精度的定位圆形标记点,提出了一种圆形标记点的识别与定位算法。在相机成像时,圆形标记点经过透视投影变换后,形态类似椭圆,用拟合椭圆的方法获取椭圆的圆心。快速获取椭圆的粗边缘点,利用高斯分布特征获取亚像素边缘点,对边缘点进行优化后利用最小二乘法拟合椭圆,可以得到最为精确的圆心坐标。实验证明,该算法用时约0.005s,椭圆圆心的位置精度在0.05pixel以内,具有很好的实用性。
In order to identify and position the center of the round marked points rapidity and high-accurately, a method is proposed. When the camera is imaging, because the round marked points are similar to the ellipse after per spective projection transformation, the center of the elliptic is obtained through elliptic fitting. The coarse edge points are got, then the sub-pixel edge points are got using the characteristic of Gaussian distribution. The edge points are optimized and fitted for ellipse by use of the method of least squares. The more accurate center coordinates are obtained. Experi ments show that the time of the algorithm is about 0. 005s and the accuracy of the ellipse＇s circle is about 0.05pixel. It is convenient for practical application.
出处
《光学技术》
CAS
CSCD
北大核心
2013年第5期393-397,共5页
Optical Technique
基金
国家自然科学基金(61171162)
教育部博士点专项基金(20110132110010)资助项目
关键词
光学测量
圆心定位
椭圆拟合
亚像素边缘点
最小二乘法
optical measurement
positioning of a center
ellipse fitting
sub-pixel edge points
least squares
