面向小直径薄壁管表面质量检测的散斑干涉图像滤波方法研究
2020-08-04陈杰阳雷罗建东
陈杰 阳雷 罗建东



摘 要
利用散斑干涉测量技术对材料表面质量进行检测时,干涉图像存在大量高频散斑噪声,严重影响了表面质量检测的准确度及测量精度。本文以小直径薄壁管表面为检测对象,提出了一种自适应窗口傅里叶变换滤波方法对散斑噪声进行滤波,首先利用小波变换获得最佳的窗口大小,然后利用窗口傅里叶脊对应的振幅动态修正频谱阈值,最后利用傅里叶脊对应的局部频率设置相应的通频带完成傅里叶逆变换获得滤波后的图像,并与等值线窗口法和二阶方向偏微分方程法进行对比分析。研究结果表明,自适应窗口傅里叶变换滤波法滤波效果最好,表面质量检测准确度及测量精度高。
关键词
小直径管;薄壁管;散斑干涉;图像滤波;自适应窗口
中图分类号: TG178 文献标识码: A
DOI:10.19694/j.cnki.issn2095-2457.2020.15.016
Abstract
The interference image contains a lot of high frequency speckle noise when using the digital speckle interferometry to measure the material surface quality, which seriously affects the accuracy of surface quality detection and measurement accuracy. In this paper, an adaptive window Fourier transform filtering method is proposed to filter the speckle noise on the surface of small-diameter thin-walled tube. Firstly, the optimal window size is obtained by wavelet transform, then the amplitude corresponding to the window Fourier ridge is used to dynamically modify the spectrum threshold, and finally the corresponding pass band is set by the local frequency corresponding to the Fourier ridge to complete the Fourier transform The filtered image is obtained by inverse transformation and compared with contoured window method and second-order directional partial differential equation method. The results show that the adaptive window Fourier transform filtering method has good filtering effect, high accuracy of surface quality detection and measurement.
Key Words
Small-diameter pipe;Thin-walled pipe;Speckle fringe pattern; Filtering;Adaptive parameters
0 引言
在核反应堆关键部件的研制过程中,必须针对小直径薄壁管的表面质量进行全面的检测及评价。相比于传统的目视、电磁、声学等检测方法,数字散斑干涉测量技术(DSPI)具有非接触、实时性、高精度及高灵敏度等特点,是实现小直径薄壁管材表面质量检测的理想技术之一[1-5]。该方法利用相干光束照射标准样品与缺陷样品表面形成反射的散斑场,散斑场会随测量对象的表面状态改变而产生相应的变化,对变化前后的散斑场进行相关运算就能得到与表面质量信息有关的干涉条纹图。然而,由于实际系统中的噪声及实验环境的影响,干涉条纹图会不可避免地存在高频散斑噪声,导致图像信噪比较低,进而直接影响相位信息提取的准确度以及最终的测量精度,因此必须有效滤除散斑干涉条纹图中的高频噪声,保证干涉测量的精度。……
