改进的盲解卷积技术在远距离激光光斑图像复原中的应用  

Application of improved blind deconvolution technology in long-ranged laser spot image restoration

作　　者：孙宏海[1,2] 刘艳滢[2] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所航空光学成像与测量中国科学院重点实验室,吉林长春130033 [2]中国科学院长春光学精密机械与物理研究所,吉林长春130033

出　　处：《中国光学》2013年第6期868-875,共8页Chinese Optics

基　　金：国家高技术研究发展计划(863计划)资助项目(No.2012AA7031010B);吉林省重大科技攻关资助项目(No.11ZDGG001)

摘　　要：为了提高外场实验中远距离测量激光光斑位置的精度,提出利用盲解卷积技术对光斑图像进行事后复原来削弱大气湍流对光斑成像的影响。首先,介绍了经典盲解卷积算法,分析了其不足之处,并提出了一种改进的盲解卷积算法。为了提高目标函数的收敛性和收敛速度,在TV(Total Variation)目标函数加入惩罚项,并对交替迭代法进行改进。然后,用数学方法证明了改进的盲解卷积算法的收敛性。最后,进行了仿真实验。与传统算法相比,用改进算法恢复的图像信噪比至少提升了15%。文中给出了外场试验图像的实际复原效果。To improve the precision of long-ranged laser spot position measurement in an outdoor experiment , blind deconvolution technology was proposed to restore the spot image afterwards to lower the impact of atmos -pheric turbulence on spot imaging .First, classical blind deconvolution algorithms were introduced , of which the shortage is analyzed .Second , an improved blind deconvolution algorithm was designed .The penalty terms were added to TV（Total Variation）objective function and the alternative iterative method was also redesigned . Then, the convergence of the improved method was testified mathematically .Finally, the improved method was compared with other classical methods by experimental simulation .The capability of improved method was testified with several objective figures , and the results show that the Signal to Noise Ratio （ SNR） is improved by 15%at least.The restoration result of outdoor experiment proves that the improved method is effective .

关 键 词：图像复原 盲解卷积 TV算法 大气湍流 激光光斑 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]