不受刀刃边缘倾角约束的遥感影像点扩散函数稳健计算方法  

A robust method of calculating point spread function from knife-edge without angular constraint in remote sensing images

作　　者：秦荣君[1] 龚健雅[1] 

机构地区：[1]武汉大学测绘遥感信息工程国家重点实验室,湖北武汉430079

出　　处：《遥感学报》2011年第5期895-907,共13页NATIONAL REMOTE SENSING BULLETIN

基　　金：国家自然科学基金创新研究群体资助项目(编号:40721001);国家重点基础研究发展计划(973计划)(编号:2006CB701304)~~

摘　　要：采用传统刀刃法提取PSF时,只有在刀刃边缘近似平行或垂直于采样方向情况下才能得到好的效果,这样限制了刀刃法提取PSF进行图像恢复的广泛应用。本文提出了一种稳健的基于倾斜刀刃的图像的PSF计算模型,不仅在提取理想刀刃状况下(与采样方向近似垂直或平行)与传统方法等价,而且对倾斜刀刃提取PSF进行图像恢复也能得到理想效果,所恢复的图像在峰值信噪比(PSNR),均方差测度(MSE)和信噪比(SNR)等方面,都要优于传统方法,利用任意倾角的刀刃边缘提取的PSF都可以基本达到理想刀刃所提取的PSF的效果,因而提高了刀刃法提取PSF进行图像恢复的鲁棒性和通用性。本文分别从理论和实践上论证了模型的有效性,并采用模拟实验数据和ADS40航空影像数据验证了方法的正确性。The traditional knife-edge method has the limitation that the chosen edge must be approximately perpendicular or parallel to the sampling direction （ideal edge）,which leads to difficulties of choosing usable knife-edge.Focusing on the issue,a new model of calculating PSF based on knife-edges with any slant angles is proposed in this paper.The proposed method is not only equivalent to the traditional one with the ideal edge,but also better than the traditional method with a slant edge in the PSNR （Peak Signal to Noise Ratio）,MSE （Mean Square Error） and SNR （Signal to Noise Ratio） of the restored image by the extracted PSF.The proposed method is robust to extract PSFs from knife-edges of any slant angles for image restoration,which extends its application.The validity of the proposed method has been reasoned both theoretically and practically,and tested with simulated data andADS40 data.

关 键 词：PSF LEVENBERG-MARQUARDT 刀刃法 维纳滤波器 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]