CCD弱目标图像条带噪声消除

光电经纬仪跟踪远距离目标时,在CCD上成像微弱。条带噪声的存在使图像质量下降,掩盖了有效的目标信息,对图像判读造成不利影响。针对条带噪声的功率谱,设计了改进的矩形低通滤波器。首先分析了条带噪声模型及其在频域的分布;然后,介绍了矩形低通滤波器通过加权的方式最大逼近真实噪声而实现噪声滤除的方法,通过实验进行矩形低通滤波器带宽的确定;最后,应用光电经纬仪CCD获得的弱目标图像进行去条带噪声实验,比较该方法与理想低通滤波及高斯带阻滤波去条带效果,结果表明该方法在有效地保留图像基本信息的前提下,获得最佳的去条带效果。

Removal of stripe noise in CCD faint targets image

The target imaging is faint on CCD , when optical-electronic theodolite traces the remote targets .Stripe noise reduces the image quality, covers useful image information and brings unfavourable effect for interpretation of images.This paper designs improved rectangular low-pass filter in view of power spectrum of striping noise .First, stripe noise model and distribution in frequency domain are analyzed .Then, rectangular low-pass filter to remove noise is introduced, which adopts the maximum approximation true noise by using the weight .The bandwidth of the rectangle low-pass filter is validated by test.Finally, the tests on faint images which are obtained by optical-electron-ic theodolite CCD show that rectangular low-pass filter reduces the stripe to a greater degree while retains the basic in-formation of image than ideal low-pass filter and gauss band stop filter.