摘要
为了提高图像定位系统的识别精度,提出了一种改进采用形态学腐蚀算法和改进的空间矩算法的十字目标亚像素中心定位方法。首先通过选取适当的结构元素,分别对组成十字图像的垂直、水平方向的直线段进行腐蚀处理,得到含有十字中心信息的水平和垂直的两条行、列像素,行、列的相交点即为十字图像特征的中点;然后选取适当参数空间,对行、列进行空间矩算法得到亚像素,最后用最小二乘法得到十字图像亚像素中心定位。实验结果表明,改进方法具有定位速度快,定位精度高的优点。
To improve the resolution and accuracy of micro - assembly systems, a new method of sub - pixel lo- calization based on Erosion operation of morphology and improved spatial moment is presented. Vertical and horizon- tal pixel lines including cross - targeting midpoint are acquired by erosion operation. The midpoint of image is the in- tersection point of vertical and horizontal pixel of result of erosion. The least square fitting edge extraction is adopted to obtain the sub - pixel localization of the cross - targeting. The experimental results indicate that the improved algo- rithm has advantages such as faster computing speed and higher location accuracy.
出处
《计算机仿真》
CSCD
北大核心
2014年第12期229-232,共4页
Computer Simulation
基金
国家自然科学基金项目(61074087)
上海市教育委员会科研创新项目(12ZZ144)
关键词
形态学
空间矩
亚像素
中心定位
Morphology
Spatial moment
Subpixel
Central location