摘要
针对微型钻针刃面视觉检测过程中的边缘角点提取问题,提出了一种基于协方差矩阵的自适应角点提取方法.通过边缘追踪算子获得微钻刃面单像素宽封闭边缘,采用自适应算法选取支持域,利用协方差矩阵获得边缘角点.应用该方法对方位各异的不同微钻进行刃面边缘角点提取,以验证该方法对不同刃面边缘角点提取的可靠性和方位无关性.
For the contour corner detection problem in the process of vision detection of micro-drill＇s first facet, an adaptive contour corner detection method based on covariance matrix was presented to detect the contour corner of micro-drill＇s first facet. One-pixel-wide closed contour of micro-drill＇s first facet is obtained by using contour-tracking operator. An adaptive arithmetic is proposed to choose the ROS （region of support） which is used in covariance matrix based contour corner detection of different micro-drill＇s first facet. The results show that the proposed method can ensure the reliability of contour corner detection of micro-drill＇s first facet.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2009年第5期825-829,833,共6页
Journal of Shanghai Jiaotong University
基金
国家自然科学基金委员会与广东省政府自然科学联合基金重点项目资助(U0734007)
关键词
微钻
自适应角点提取
协方差矩阵
边缘追踪
micro-drill
adaptive corner detection
covariance matrix
contour-tracking
