面向图像复原的广义岭估计模式法波前重构  

Wavefront reconstruction for image restoration based on generalized ridge estimation

作　　者：陈波[1] 程承旗[1] 濮国梁[1] 耿则勋[2] 杨阳[2] 

机构地区：[1]北京大学工学院,北京100871 [2]信息工程大学测绘学院,河南郑州450052

出　　处：《红外与激光工程》2012年第2期478-483,共6页Infrared and Laser Engineering

基　　金：国家自然科学基金(41001237);中国博士后科学基金面上资助项目(20100470110);中国博士后科学基金特别资助项目(201104001)

摘　　要：为了提高自适应光学图像的复原效果,提出了一种面向图像复原的广义岭估计Zernike模式波前重构算法。首先,按照用时间换精度的思路,把广义岭估计引入自适应光学波前重构算法中。改善最小二乘估计的奇异性,抑制波前测量误差的放大。然后,引入迭代计算提高方程的解算精度,进一步提高点扩散函数(PSF)重建质量。最后,利用广义岭估计重建的PSF对降质图像进行复原实验验证PSF的重建质量。实验结果证明,广义岭估计PSF的复原效果比最小二乘估计PSF高约15%,更接近成像系统的光学衍射极,同时复原计算迭代次数也减少了35%。新算法重构的PSF在图像复原中有更好的图像重建效果。In order to enhance the performance of adaptive optics image restoration,a novel wavefront reconstruction algorithm based on generalized ridge estimation was proposed.In the view of adaptive optics image post-processed restoration,not real time and fast computation but the high-precision was the basic request of wavefront reconstruction.On the consideration of sacrifice time for precision,it brought the idea of generalized ridge estimation（GRE） into wavefront-reconstruction algorithm,to improve the singularity of least square（LS） and restrain the enlarger of wavefront measure-error.Then,the iterative solution method was brought into wavefront-reconstruction for improving the solution of singular equation and the precision of reconstructed-point spread function（PSF）.Finally,the blurred-image restoration experiment was implemented to verify the precision of reconstructed-PSF.The experiment proves that the novel algorithm which restores PSF performes well in image restoration,and the image restoration effect of GRE-Reconstructed-PSF was 15% better than those of LS-Reconstructed-PSF.

关 键 词：图像复原 自适应光学 波前重构 广义岭估计 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]