帧转移型CCD传感器在拍摄星空背景图像时，视场中高亮度恒星会产生Smear现象，从而对目标检测有严重影响.为了消除Smear现象，提出了一种星图中基于小波变换的Smear消除方法.首先根据Smear产生机理及星空图像特性，建立星图Smear问题的模型；然后使用多层二维Haar小波分解，把Smear亮线分离到低频分量及高频垂直分量中分别进行Smear消除处理；最后重构出消除Smear的图像.实验结果表明：该方法能有效去除Smear，最大限度保留图像原有信息，并可增强Smear区域弱小目标信噪比.

When the frame transfer CCD image sensor shooting starsky background image,the smear phenomenon caused by high brightness stars in the field of view will seriously affect the target detection.For removing the smear phenomenon,a smear removal algorithm based on wavelet transform in starsky image was proposed.According to the mechanism of smear and characteristics of the starsky image,the model of smear problem was established,the smear line was separated into the high frequency components and vertical components by using multilayer twodimension haar wavelet transform,and the smearremoved image was reconstructed.The expirmental results show that the method can effectively remove smear and retain original image information,and enhance the SNR of small target in smear area.