气动退化图像的改进复原算法  

Improved Restoration Algorithm of Aero Degraded Image

作　　者：孙李辉 李俊山[1] 李钊[2] 韩洪伟[3] 胡双演[1] 

机构地区：[1]第二炮兵工程大学 [2]二炮驻石家庄地区军事代表室 [3]中国人民解放军96656部队

出　　处：《无线电工程》2012年第10期43-44,54,共3页Radio Engineering

基　　金：国家自然科学基金项目(61175120)

摘　　要：针对气动光学效应引起的图像模糊退化导致复原处理困难的问题,提出了一种气动退化红外图像的改进复原算法。具体做法是:在假设退化图像的点扩散函数服从高斯分布的前提下,对图像进行带参数维纳滤波,在此过程中,使用图像清晰度评价函数作为复原效果的评价实现了算法中参数的最优化。通过多组仿真实验和算法的对比分析表明,算法较好地保持了图像细节,较好地恢复了图像质量,实时性较好。To address the difficulty in restoration of aero degraded infrared image due to aero-optical effects, an improved restora- tion algorithm of aero degraded infrared image is proposed. First, wiener filtering is conducted assuming that the point spread function （PSF） is subject to Gaussian distribution. During this process, image definition evaluation function is used for evaluation of restoration effect to achieve optimization. Numerical experiment results and comparisons show that this algorithm is better in detail-preserving and deblurring of aero degraded infrared image. The real time performance is also better.

关 键 词：气动光学效应 复原 点扩散函数 图像清晰度评价函数 实时 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]