Abstract: In order to improve the resolution and accuracy of micro-assembly systems，a novel sub-pixel edge detection and center localization algorithm for micro-parts is presented.Firstly，the invariant moments are employed to describe the feature of the micro parts.Then，the edge of micro parts at pixel level is extracted using Canny operator.Secondly，Zernike orthogonal moments are used to relocate the edge of microsize part at sub-pixel level.Lastly，least squares ellipse fitting is applied to position the center.Experiment results show that the algorithm can achieve higher location accuracy and consume less time.

Key words: subpixel edge extraction, Zernike moments, least squares ellipse fitting, center location, micromanipulation

摘要： 为了提高微操作系统的装配精度，提出了一种新型的亚像素边缘检测和中心定位算法。应用Canny算子提取了微零件在像素级的边缘。应用Zernike矩对微零件进行亚像素级的边缘定位。采用最小二乘椭圆拟合定位微零件的中心位置。实验结果表明，该算法能够实现更高的定位精度和消耗更少的时间。

关键词: 亚像素边缘提取, Zernike矩, 最小二乘椭圆拟和, 中心定位, 微操作