基于自适应拟合的小型元件位姿高精度检测算法
|
摘要
针对小型元件在生产过程中精度要求高的难点,以一种典型的小型元件,发火元件的自动化焊接流程为例,对流程中最重要的视觉检测步骤,提出了发火元件的亚像素轮廓自适应拟合算法。在算法中,首先利用感兴趣区域提取、高斯滤波、Otsu阈值分割、Freeman轮廓跟踪,进行图像预处理;其次利用Zernike矩对跟踪编码的轮廓进行亚像素化,提高轮廓检测精度;最后根据亚像素化后的轮廓,进行自适应轮廓拟合,从而实现发火元件位姿的自动检测。对上述算法,在发火元件测试平台上验证,实验得到最大定位误差9.37μm,平均误差5.96μm;转角误差0.006°,满足焊接检测的要求。
To deal with the problems of micro components in high welding precision during the production process, a typical kind of micro component, ignition component is taken as example, study on machine vision is carried out to realize the welding process, and subpixel contour adaptive fitting algorithm for ignition components is proposed. In the algorithm, firstly, region of interest extraction, Gauss filtering, Otsu threshold segmentation and Freeman contour tracking are used for image preprocessing. Secondly, the Zernike moments are used to transfer the contour of the tracking code into subpixel level to improve the accuracy of contour detection. Finally, adaptive contour fitting is implemented based on the subpixel contour to automatically detect the pose of ignition components. The above algorithm is verified on the ignition component test platform, the maximum positioning error is 9.37 microns, average error is 5.96 microns, and the angle error is 0.006 degrees, which meets the requirements of the welding detection.
To deal with the problems of micro components in high welding precision during the production process, a typical kind of micro component, ignition component is taken as example, study on machine vision is carried out to realize the welding process, and subpixel contour adaptive fitting algorithm for ignition components is proposed. In the algorithm, firstly, region of interest extraction, Gauss filtering, Otsu threshold segmentation and Freeman contour tracking are used for image preprocessing. Secondly, the Zernike moments are used to transfer the contour of the tracking code into subpixel level to improve the accuracy of contour detection. Finally, adaptive contour fitting is implemented based on the subpixel contour to automatically detect the pose of ignition components. The above algorithm is verified on the ignition component test platform, the maximum positioning error is 9.37 microns, average error is 5.96 microns, and the angle error is 0.006 degrees, which meets the requirements of the welding detection.
引文
[1]TRUJILLO-PINO A,KRISSIAN K,ALEMN-FLORES M,et al.Accurate subpixel edge location based on partial area effect[J].Image and Vision Computing,2013,31(1):72-90.
[2]SUN Q C,HOU Y Q,TAN Q C.A subpixel edge detection method based on an arctangent edge mode[J].Optik-International Journal for Light and Electron Optics,2016,127(14):5702-5710.
[3]MAISELIA B J,GAO H J.Robust edge detector based on anisotropic diffusion-driven process[J].Information Processing Letters,2016,116(5):373-378.
[4]KALRA A,CHHOKAR R L.A hybrid approach using sobel and canny operator for digital image edge detection[C]//International Conference on Micro-Electronics and Telecommunication Engineering.IEEE,2017:305-310.
[5]朱训生,任新生,薛秉源.缺口圆截面的圆度评定及其最小二乘圆度公式[J].机械设计与研究,1994,10(3):38-39.
[6]CHERNOV N,LESORT C.Least squares fitting of circles[J].Journal of Mathematical Imaging and Vision,2005,23(3):239-252.
[7]刘珂,周富强,张广军.半径约束最小二乘圆拟合方法及其误差分析[J].光电子·激光,2006,17(5):604-607.
[8]LIN S,JUSKO O,HRTIG F,et al.A least squares algorithm for fitting data points to a circular arc cam[J].Measurement,2017,102(5):170-178.
[9]FREEMAN H,DAVIS L S.A corner finding algorithm for chain code curves[J].IEEE Tran.Compt.,1977,26(3):297-303.
[10]SONKA M,HLAVAC V,BOYLE R.Image processing,Analysis and Machine vision[M].4rth edition,Beijing:Tsinghua University Press,2016:97-106.
[11]金守峰,范荻,陈蓉,等.面向回转类零件圆度的机器视觉测量方法与试验[J].机械设计与研究,2016,32(4):117-119.
[12]徐要刚,赵现朝,田志松.激光测距法测量大锻件尺寸[J].机械设计与研究,2008,24(6):94-97.
[13]GHOSAL S,MEHROTRA R.Orthogonal moment operators for subpixel edge detection[J].Pattern Recognition,1993,26(2):295-306.
[14]ZHANG Z Y.Aflexibe new technique for camera calibration[J].IEEE Trans.on Pattern Analysis and Machine Intelligence.2000,22(11):1330-1334.
[15]方素平,夏晓华,肖燕,等.一种线阵相机镜头畸变的标定方法[J].西安交通大学学报,2013,47(1):11-14.
[2]SUN Q C,HOU Y Q,TAN Q C.A subpixel edge detection method based on an arctangent edge mode[J].Optik-International Journal for Light and Electron Optics,2016,127(14):5702-5710.
[3]MAISELIA B J,GAO H J.Robust edge detector based on anisotropic diffusion-driven process[J].Information Processing Letters,2016,116(5):373-378.
[4]KALRA A,CHHOKAR R L.A hybrid approach using sobel and canny operator for digital image edge detection[C]//International Conference on Micro-Electronics and Telecommunication Engineering.IEEE,2017:305-310.
[5]朱训生,任新生,薛秉源.缺口圆截面的圆度评定及其最小二乘圆度公式[J].机械设计与研究,1994,10(3):38-39.
[6]CHERNOV N,LESORT C.Least squares fitting of circles[J].Journal of Mathematical Imaging and Vision,2005,23(3):239-252.
[7]刘珂,周富强,张广军.半径约束最小二乘圆拟合方法及其误差分析[J].光电子·激光,2006,17(5):604-607.
[8]LIN S,JUSKO O,HRTIG F,et al.A least squares algorithm for fitting data points to a circular arc cam[J].Measurement,2017,102(5):170-178.
[9]FREEMAN H,DAVIS L S.A corner finding algorithm for chain code curves[J].IEEE Tran.Compt.,1977,26(3):297-303.
[10]SONKA M,HLAVAC V,BOYLE R.Image processing,Analysis and Machine vision[M].4rth edition,Beijing:Tsinghua University Press,2016:97-106.
[11]金守峰,范荻,陈蓉,等.面向回转类零件圆度的机器视觉测量方法与试验[J].机械设计与研究,2016,32(4):117-119.
[12]徐要刚,赵现朝,田志松.激光测距法测量大锻件尺寸[J].机械设计与研究,2008,24(6):94-97.
[13]GHOSAL S,MEHROTRA R.Orthogonal moment operators for subpixel edge detection[J].Pattern Recognition,1993,26(2):295-306.
[14]ZHANG Z Y.Aflexibe new technique for camera calibration[J].IEEE Trans.on Pattern Analysis and Machine Intelligence.2000,22(11):1330-1334.
[15]方素平,夏晓华,肖燕,等.一种线阵相机镜头畸变的标定方法[J].西安交通大学学报,2013,47(1):11-14.