基于小波变换的RL湍流退化图像复原算法  

RL turbulence degraded image restoration algorithm based on wavelet transform

作　　者：徐晓睿 戴明[1] 尹传历[1] XU Xiao-rui DAI Ming YIN Chuan-li(Changchun Institute of Optics, Fine Mechanics and Physics, Cinese Academy of Sciences, Changchun 130033 ,China University of Chinese Academy of Sciences, Beijing 100049 ,China)

机构地区：[1]中国科学院长春光学精密机械与物理研究所,长春130033 [2]中国科学院大学,北京100049

出　　处：《液晶与显示》2017年第10期822-827,共6页Chinese Journal of Liquid Crystals and Displays

基　　金：国家自然基金(No.61405191)~~

摘　　要：为了从湍流退化图像中准确有效地恢复出目标图像,提出一种基于小波变换的RL湍流退化图像复原算法。该算法首先对湍流退化图像进行小波分解,可得到不同分解尺度下,不同频带的子图像。根据不同方向的高频子段的小波系数,估计出各个高频子段噪声方差,进而求得适用于各频段的自适应阈值,以这些阈值为软阈值法的临界条件分别对各频段的小波系数进行收缩,最后用RL算法去迭代小波重构后的图像来实现湍流退化图像的复原。为了验证该方法的有效性,分别用这两种算法在不同噪声条件下,对同一幅退化图像进行了仿真实验。改进后的算法使得两幅图像的峰值信噪比分别提高5.894 3dB和7.108 4dB。结果表明,本文的算法相比RL算法在复原效果上有一定的提高。In order to recover the target image accurately and effectively from the turbulence degraded image, a RL turbulence degraded image restoration algorithm based on wavelet transform is proposed. Firstly, the turbulence degraded image is decomposed by wavelet, and sub-images of different fre- quency bands can be obtained at different decomposition scales. According to the wavelet coefficients of the high frequency sub-segments in different directions, the noise variance of each high frequency sub-segment is estimated, and then the adaptive thresholds suitable for each frequency band are ob- tained. The critical conditions of these thresholds are the wavelet. And then the RL algorithm is used to iterate the reconstructed image to reconstruct the turbulence image. In order to verify the effective- ness of the method, the simulation results of the same degraded image are simulated by the two algo- rithms under different noise conditions. The improved method makes the peak signal to noise ratio of the two images increase by 5.894 3 dB and 7.108 4 dB. The results show that the proposed algorithm has a certain improvement in the recovery effect compared with the RL algorithm.

关 键 词：Richardson-Lucy算法 小波分解 小波去噪 软阈值法 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]