摘要
提出了一种高精度零件尺寸测量的亚像素边缘检测算法,采用改进的Sobel算子去掉局部非极大值点获得像素级边缘,在梯度方向上进行高斯曲线拟合,可实现图像边缘亚像素精确定位,提高测量系统的精度。
A subpixel edge detecting algorithm used in the highly accuracy measurement of parts is introduced. At first one pixel edge is obtained by getting rid of these non-maximum points in local area based on Sobel operator, and the precision of image edge location is more improved through Gauss curve fitting and interpolation in the gradient direction, consequently the drecision of measurement system is highly improved.
出处
《重庆科技学院学报(自然科学版)》
CAS
2008年第6期82-84,共3页
Journal of Chongqing University of Science and Technology:Natural Sciences Edition
基金
湖南省教育厅科研基金资助项目(07D005)
关键词
图像测量
亚像素
曲线拟合
边缘检测
imaging measurement
subpixel
curve fitting
edge detection