线阵CCD卫星图像自适应条带噪声去除

摘要

线阵CCD(charge coupled device)卫星原始数据普遍存在由于传感器响应不均匀或CCD拼接所造成的周期性条带噪声,降低了卫星图像的质量和应用效果.在分析线阵CCD卫星图像噪声特性的基础上,采用基于灰度统计的自适应空间滤波方法,进行基于BJ-1卫星和CBERS-02卫星原始图像的条带噪声去除实验,并与直方图匹配、频域滤波等去条带噪声算法相比较.实验结果表明,此方法可有效去除列条带噪声,并均衡CCD响应差异,从而改善图像质量.%Periodic stripe noise is an obvious phenomenon in linear-array CCD satellite raw image because of response difference and stitching among CCD detectors,which reduces image quality and application effects.On the basis of analysis of noise characteristics within linear-array CCD satellite image,this paper adopts adaptive spatial filtering algorithm based on gray statistic to remove stripe noise in BJ-1 satellite raw image and CBERS-02 satellite raw image.After comparison to the processing results of histogram matching de-striping algorithm and frequency filtering de-striping algorithm,the results show that this algorithm can effectively demolish vertical stripes in ALOS satellite imagery and balance response difference among CCD detectors to improve the images quality.