供稿: 王小林;熊新海;张高峰 | 时间: 2019-05-05 | 次数: |
作者单位:河南理工大学机械与动力工程学院
摘要:在对数字图像中的螺旋锥齿轮进行空间定位和三维重构过程中,为获取齿轮轴向参数,提出了最小二乘法与圆度轮廓误差评定原理;建立了锥齿轮数字图像仿真模型.来获取齿轮轴向参数;利用Matlab软件对数字图像进行图像增强、二值化、边界提取等操作,再结合最小二乘法对像素坐标进行椭圆拟合;通过利用圆度轮廓误差评定的方法,使拟合获得的椭圆误差最小.该方法能准确地拟合出椭圆,通过轮廓逼近算法有效地去除了噪声点,提高了椭圆拟合的精度,可为螺旋锥齿轮空间定位和三维重构提供准确的数据.
基金:河南省科技攻关项目(072102210033);河南理工大学博士基金资助项目(B050701);
DOI:10.16186/j.cnki.1673-9787.2008.03.017
分类号:TP391.41
The method of ellipse contour extraction based on Least-squares and elliptic roundness error
Abstract:In order to obtain gear axial parameters from the spiral bevel gear space positioning and three-dimensional reconstruction of digital image, the principle of the least squares, the round contour error evaluation is proposed.The simulation model of the bevel gear digital image is estascished for access to the gear axial parameters.Matlab software is used for digital images processing, such as image enhancement, the border extraction, the border value, combined with the least squares coordinates of pixels to fit oval, and to make the oval fitting error minimum by using of the round contour error evaluation.The method can fit ellipse accurately, get more accurate parameters of the oval and effectively remove the noise, improve the accuracy of fitting oval, provide a algorithm of the digital image to remove noise, and offer the more accurate data of the spiral bevel gear for positioning and three-dimensional reconstruction.