摘要
介绍一种基于CCD的光学自准直系统观测方法及相关的数字图像处理方法。关于图像处理,首先采用自动阈值分割提取十字丝的像素级边缘,然后使用积分法找到十字丝的像素级边缘位置,接着运用三次多项式拟合出图像边缘实现十字丝边缘的亚像素定位,最后对十字丝亚像素边缘进行直线拟合得到边缘直线方程,进而得到十字丝中心。实验结果表明,亚像素细分技术能够较好地提取十字丝亚像素边缘。
An observation method of optical auto-collimation system based on CCD is proposed and the image processing method is introduced. At first, auto-threshold segmentation is used to detect the pixel edge of reticle. Then, the integral method is used to locate the pixel edge of reticle. The following processing is cubic polynomial fit- ting which fit the image edge to implement sub-pixel localization. Finally, straight line fitting is applied and the cen- ter of reticle is obtained. The experiment results show that the sub-pixel division can detect the sub-pixel edge of reticle precisely.
出处
《大地测量与地球动力学》
CSCD
北大核心
2013年第A02期100-102,105,共4页
Journal of Geodesy and Geodynamics
基金
地震科技星火计划青年项目(XH2033Y)
关键词
边缘提取
亚像素细分
自动阈值分割
积分法
多项式拟合
edge detection
sub-pixel division
auto-threshold segmentation
integral method
polynomial fitting
