小卫星CCD相机MTF在轨测量与图像复原

On-orbit MTF estimation and restoration for CCD cameras of environment and disaster reduction small satellites(HJ1A/1B)

基于调制传递函数的基本理论,针对环境与减灾小卫星光学相机,基于刃边法给出了一种在轨测量成像系统调制传递函数的方法.通过选取图像中均匀地物之间灰度变化明显的边缘,对有限采样点进行平滑滤波、曲线拟合和对线扩展函数(LSF)的傅里叶变换获得点扩展函数(PSF),提取调制传递函数(MTF)曲线,进而选取最优的MTF曲线.考虑到采用有约束复原方法时噪声统计、估计的复杂性和欠普适性,本方法通过频率域的去卷积进行复原处理,其优势在于不需要提供传感器、大气状况等参数.并对复原图像从统计特征、清晰度、纹理特征等3方面进行了定量分析和评价.结果表明:复原后的环境与减灾小卫星图像质量得到明显提高,收到良好的复原效果.

Focusing on the CCD cameras of environment and disaster reduction small satellites（HJ1A/1B）,an approach on the on-orbit measurement of modulation transfer function（MTF） to the imaging system was proposed.According to selecting the distinct edges among two homogeneous objects from images,the degraded images were processed with smooth filtering,curve fitting and the forward Fourier transform（FFT） of line spread function（LSF） to obtain the point spread function（PSF）.After that,MTF corresponding to PSF can be obtained through FFT.Taking into account the noise of the complex constraint restoration method（CRM） without universality,the deconvolution restoration method in frequency domain was proposed,because it had the advantage that we neet not prove the parameters of sensor and atmosphere.Finally,analyzing and evaluating the restored image were carried out based on three factors： the feature value,definition and texture,respectively.The results show that the quality of HJ1A/1B image has been obviously improved,and achieved good restoration results.