基于傅里叶变换的条带噪声去除方法研究  

Study on Removing the Stripe Noises based on Fourier Transform

作　　者：杜艺[1] 刘文国[1] 葛帅 

机构地区：[1]山东省水利勘测设计院,山东济南250013 [2]济南市房产测绘研究院,山东济南250001

出　　处：《测绘与空间地理信息》2014年第8期165-167,共3页Geomatics & Spatial Information Technology

摘　　要：提出一种新的基于傅里叶变换去除条带噪声的方法,首先将图像从空间域转换到频率域,通过频谱图发现条纹噪声的周期性分布,根据傅里叶变换的线性性质、周期性质和图像在频率域越高频分量处频谱值越小的性质,去除条带噪声,与基于低通滤波的条带噪声去除方法进行比较,可以发现本文提出的方法效果较好,为遥感影像的条纹噪声去除方法研究提供了一个参考。The paper presents a new method of removing stripe noises based on fourier transform , the image is firstly transferred from space domain to frequency domain , by observing the period distribution of the stripe noises on the spectrogram plot , utilizing the prop-erties of linear and periodicity and the fact that the magnitude of Fourier Transform gets smaller for higher frequencies which is farther from the center on the spectral graph , the stripe noises are removed .Compared to the traditional mean , it shows that the result is bet-ter, the method could be theoretical reference for removing stripe noises in the remote sensing images .

关 键 词：条带噪声 傅里叶变换 低通滤波 

分 类 号：P2[天文地球—测绘科学与技术]