基于FFT的条纹图规则噪声滤波  

Filter of Regular Noises for Fringe Pattern Based on FFT

作　　者：张伟[1] 朱福龙[2] 

机构地区：[1]湖北汽车工业学院机械工程学院,湖北十堰442002 [2]华中科技大学机械科学与工程学院,湖北武汉430074

出　　处：《湖北汽车工业学院学报》2013年第3期61-64,共4页Journal of Hubei University Of Automotive Technology

基　　金：国家科技重大专项资助项目(2009ZX02010-041)

摘　　要：影子云纹测量技术用于微电子封装基板翘曲测量中,条纹图往往受到被测基板均布芯片产生的规则噪声的影响,导致包裹相位质量较差,不能正确解包,引起较大的测量误差。针对条纹图中的规则噪声,分别采用常用的均值滤波、中值滤波和傅里叶变换滤波方法处理条纹图,傅里叶变换滤波方法能有效消除条纹图中规则背景噪声的影响,改善包裹相位图质量,提高测量精度。Shadow moire microelectronics packaging, technique is Widely Used in measuring the warpage of substrate of the fringe pattern is influenced by the regular noises caused by the uniformly assembling chips. The noises cause the lower quality of wrappedphase and cannot correctly perform phase unwrapping. For denoising the regular noises of fringe pattern, the mean filter, median fiher and Fast Fourier Transform filter were adopted to improve the quality of fringe pattern. The image processing results show that the FFT filter can decrease the influence of regular noises and increase the measurement precision. Key words： fringe pattern; regular noise; filter; Fast Fourier Transform （FFT）

关 键 词：条纹图 规则噪声 滤波 快速傅里叶变换 

分 类 号：TH741[机械工程—光学工程]