离焦模糊图像超分辨力盲复原算法分析  

Analysis on super-resolution blind restoration algorithm of defocus blurred image

作　　者：赵琳[1] 金伟其[1] 黄有为[1] 王岭雪[1] 

机构地区：[1]北京理工大学信息科学技术学院光电工程系,北京100081

出　　处：《光学技术》2010年第1期75-78,共4页Optical Technique

基　　金：中国信息安全产品测评认证中心开放基金资助项目(P2006019EF)

摘　　要：根据计算机模拟和实际拍摄的离焦模糊标准分辨力靶标图像,对提出的离焦模糊图像超分辨力盲复原算法进行了分析研究,对比了超分辨力盲复原算法与几种典型复原算法的效果,以及在不同离焦模糊半径下该算法的复原效果。实验结果表明,该算法对深度离焦模糊图像具有较明显的复原效果,复原图像相比较于模糊图像分辨力提高1倍;显著改善了标准分辨力靶标图像的可分辨组块。可分辨组块可以定量地描述复原算法对图像分辨力的改善程度。该算法已经应用到实际工作当中,并获得了较好的效果。Analyses are developed on super-resolution blind restoration algorithm of defocus blurred image by utilizing the simulation defocus blurred resolving power image and captured practice ones.Comparison is carried through between the super-resolution and other algorithms.Different effects also are compared when the radius of defocus blur changes.The result indicates that this algorithm can obtain good effect when the defocus blur is severely and the resolving power of the restored image is improved one times compared with the blurred case.The visible resolving power diamonds of the standard drone image are increased prominently.The visible resolving power diamonds can just be used to describe the improvement of the resolving power of the image.The proposed method is successfully applied into actual metric and received better results.

关 键 词：图像盲复原 离焦模糊 分辨力靶标 超分辨力 

分 类 号：TN911.73[电子电信—通信与信息系统]