摘要
针对机器视觉测量中图像数据处理时海量数据处理和剔除误差点等问题,提出了一套合理、有效的误差点剔除方法。构建了一个基于机器视觉的精密检测平台,并根据测量中的目标边缘,针对最小二乘直线拟合和圆弧拟合算法进行了实验分析,得到了测量系统的测量精度:尺寸测量精度约为5μ,角度精度约为2'。在很大程度上提高了视觉测量系统的测量精度与测量效率,是机器视觉在精密检测中的一次成功应用。
A reasonable and effective method of deleting error points is proposed in view of the problem of massive information and error point deletion when processing images in machine vision measurement. The article constructs a precision detection platform based on machine vision, and based on the measurement edge, this article analyzes it with the least squares line fitting and circular arc fitting. The position precision reaches 5 μ and the angle measuring precise reaches 2′, which improves the detection precise and detection efficiency and is a successful application in precision detection.
出处
《重庆理工大学学报（自然科学）》
CAS
2010年第3期66-70,共5页
Journal of Chongqing University of Technology：Natural Science
基金
重庆市自然科学基金资助项目(CSTS
2006BB3176)
关键词
图像处理
图像分析
轮廓描述
精密检测
image processing
image analysis
outline description
precision detection
