首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于旋滤波法的干涉条纹预处理技术
引用本文:徐建程,许乔,柴立群,邓燕.基于旋滤波法的干涉条纹预处理技术[J].强激光与粒子束,2006,18(1):69-72.
作者姓名:徐建程  许乔  柴立群  邓燕
作者单位:成都精密光学工程研究中心, 成都 610041
基金项目:中国工程物理研究院基金资助课题
摘    要: 提出了一种新的基于旋滤波法的干涉条纹预处理方法。该方法根据条纹图灰度值梯度分布规律,只在干涉条纹切线方向进行中值滤波,它能有效地处理各种相干噪声而不会使条纹变模糊。实验证明该方法能有效地去除随机噪声、椒盐噪声以及由于光照不均匀引起的较大面积的亮斑或暗斑等噪声。对于亮斑或暗斑,其滤波效果要明显优于Gerchberg外插迭代算法。以WYKO相移干涉仪5次测量的面形平均值作为被测面形的参考值,则经过预处理后,抑制了噪声引起的局部误差,傅里叶变换法计算得到PV值的误差从未预处理的14.4%(全口径)、131%(95%口径)分别减小到3.4%(全口径)、1.5%(95%口径);RMS误差从28.7%(全口径)、23.3%(95%口径)分别减小到5.0%(全口径)、2.3%(95%口径)。

关 键 词:光学检测  干涉测量  干涉条纹分析  傅里叶变换  中值旋滤波法
文章编号:1001-4322(2006)01-0069-04
收稿时间:2005-06-16
修稿时间:2005-10-25

A new fringe-pattern preprocess method based on median spin filtering
XU Jian-cheng,XU Qiao,CHAI Li-qun,DENG Yan.A new fringe-pattern preprocess method based on median spin filtering[J].High Power Laser and Particle Beams,2006,18(1):69-72.
Authors:XU Jian-cheng  XU Qiao  CHAI Li-qun  DENG Yan
Institution:Fine Optical Engineering Research Center, P.O.Box 450, Chengdu 610041, China
Abstract:A new fringe-pattern preprocess based on median spin filter is proposed to deal with the interferogram with high-level noises.It only smoothes noises in fringe-tangential direction by median filter,therefore,it can filter off noises such as coherent artifacts,additive random noise,"Salt and pepper" and large bright and black spots,without blurring and distorting the fringe-pattern features.The experimental results show that the median spin filter is suitable for the interferogram with all kinds of noise,especially the random noise,speckle noise and large bright or black spots.The method is better than Gerchberg's extrapolation algorithm and filters in frequency space when dealing with large bright or black spots.The P-V and RMS errors are less than 1.5% and 2.3% for 95% aperture by this method.
Keywords:Optical test  Interferometry  Fourier transform  Fringe-pattern analysis  Median spin filtering
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《强激光与粒子束》浏览原始摘要信息
点击此处可从《强激光与粒子束》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号