摘要
在线相位测量轮廓术(PMP)中,当被测物体运动速度较高时,所采集的变形条纹往往为运动模糊图像,使得复原误差增大,严重时可能导致三维重建无法进行。将在线PMP运用于钢轨外形及表面缺陷的在线三维测量时,为了实现钢轨表面模糊变形条纹的清晰化,本文对维纳滤波法、点扩散函数算法、盲解卷积算法和Richardson-Lucy算法等几种模糊图像复原算法进行了对比分析,用峰值信噪比对模糊条纹图像的复原效果进行评估。同时,研究了车辆运行速度和图像复原效果之间的关系,得出了复原效果与运行速度之间的关系曲线,进行了误差分析,并用在线PMP实现了钢轨外形的三维重建。理论及实验结果表明:在对钢轨外形轮廓及表面缺陷的在线三维测量时,Richardson-Lucy算法的复原效果最佳,图像复原程度与车辆运行速度呈多项式关系。
In online phase measurement profilometry（PMP）,when the which velocity of the the shape object is quite high,error,flaw,the or collected even reformed to fringes often appear to be motion-blurred the images,increases measurement and leads clarify the the 3D reconstruction failure.In online rail,PMP 3D measurement of rail such surface in order to blurred reformed fringes of the several restoration methods,as Wiener were Filtering,compared Point and Spread Function And algorithm,peak Blind to Deconvolution noise algorithm,and for Richardson-Lucy evaluating algorithm analyzed.relationship shape the signal vehicle ratio and（PSNR）is used the the restoration error effect.Meanwhile,the between the speed on the PMP.restoration effect was studied,was show analyzed that and the 3D rail was reconstructed rail based online flaw,Theoretical and experimental algorithm results best in the online 3D of measurement motion of and shape and surface the Richardson-Lucy is the the one for image is restoration blur,the relationship between the effect of image restoration and vehicle speed a polynomial.
出处
《光电工程》
CAS
CSCD
北大核心
2017年第7期695-700,共6页
Opto-Electronic Engineering
基金
国家自然科学基金资助项目(61471304)
作者简介
E—mail：jinlong—lee@126.com
