小波分析在中巴资源一号卫星图像预处理中的应用  

A WAVELET BASED METHOD OF SMOOTHING STRIP INTERFERE FROM SATELLITE IMAGE

作　　者：陈付华[1] 屈颖歌[1] 夏德深[1] 

机构地区：[1]南京理工大学计算机系,南京210094

出　　处：《计算机应用与软件》2003年第9期66-68,共3页Computer Applications and Software

摘　　要：中巴资源一号卫星图像中存在比较明显的条带干扰现象。本文利用小波分析的多分辨和局部化性质 ,根据卫星探测头中探测元的个数 ,精确地选择小波分解中相应的子图进行处理 ,针对性强 ,效果理想。在处理拖尾现象和保持图像信息方面 ,该方法优于基于Fourier分析的频域滤波方法。The usual method used to smooth strip is filtering in frequency field.Using the multi-scale and localization properties of wavelet,and according to the number of detect-elements of a detector in satellite,an proper sub-image which has good pertinence can be got by wavelet decompose.This method is better than Fourier transform based filtering in frequency field.Experiments with CBERS-1 image have shown that this method is of better result and higher speed.

关 键 词：中巴资源一号卫星图像 预处理 小波分析 卫星探测头 频域滤波 条带干扰 

分 类 号：V474[航空宇航科学与技术—飞行器设计]