摘要
利用Talbot-Moiré技术测量长焦距时,条纹中心的定位精度直接影响条纹宽度或斜率的计算精度.提出采用Canny算子和形态学方法对条纹中心进行像素级粗定位,用高斯曲线对条纹灰度进行拟合对其进行亚像素定位,并用组合测量的方法和最小二乘法来处理数据,减小了随机误差对条纹中心定位精度的影响.经过对光强符合正弦分布的标准条纹的定位,验证了采用本文的定位方法可以减小条纹中心定位误差,从而提高条纹周期的计算精度.并实际测量了光栅的栅距,实验结果再次证明采用组合测量的数据处理的方法可以提高亚像素定位精度.
As the focus of research,measurement based on Talbot effect and Moiréfringes is suitable to long focal length of lens.To get the focal length accurately,the width or slope coefficients of Moiré fringes are essential parameters.So the center positioning precision must be improved due to its significant impact on accuracy of the width or slope coefficients of fringes.A method of center sub pixel localization is put forward.Canny operator and detection based on mathematical morphology are applied to determine the location of fringes centre at the level of pixel.The sub pixel localization is presented based on Gauss curve fitting.To reduce the influence of error,the combination measurement and least square method are used in data processing.Finally,the localization error is proved to be reduced by positioning the standard fringes generated.The result of gratings measurement shows that the combination measurement can improve the precision.
出处
《西安工业大学学报》
CAS
2013年第4期289-292,308,共5页
Journal of Xi’an Technological University
关键词
图像处理
莫尔条纹
亚像素
中心定位
image processing
moiréfringes
sub pixel
center localization
