摘要
在采用图像检测技术对轴套类零件尺寸进行高精度检测中,提出了一种高精度零件尺寸图像检测方法,在获取零件图像后,采用Prewitt算子完成对图像边缘的初步定位,在此基础上,通过对图像边缘灰度变化的离散值作最小二乘曲线拟合,并对该拟合曲线求极值,得到边缘位置;为了减少干扰对测量值的影响,采用误差数据处理方法筛选出有一定精度的检测数据,然后对这些检测数据求平均值,获得精确的边缘位置,由此计算出精确的零件尺寸。实验表明,该方法能有效地提高检测精度,在检测分辨率为1000DPI情况下,检测精度可达到13μm。
In order to gain the precise size of the axis parts by optoelectronic detecting technology, a high precision method of image measuring was presented. The image of part by Prewitt operator was located. The dispersed data of intensity gradient were processed by the method of least-squares curve-fitting for detecting the edge of image. The measurement of image could be calculated. In order to decrease the influence of the interfere, the error theory was used to select the correct data. The accurate position of the edge was obtained by the average of the data The accurate size of the part was calculated. The experimental results Drove that the error is 13 μm when the resolution is 1000 DPI.
出处
《光学与光电技术》
2008年第2期85-87,共3页
Optics & Optoelectronic Technology
基金
广东省教育厅科研(05Z010)资助项目
关键词
光电检测
图像检测
边缘
最小二乘曲线拟合
误差理论
optoelectronic detecting
image measuring
edge
least-squares curve-fitting
error theory
