RAISR算法在遥感图像超分辨率重建中的可行性  

Feasibility of RAISR Algorithm in Remote Sensing Image Super Resolution Reconstruction

作　　者：卜丽静[1] 吴文玉 张正鹏[1] BU Lijing;WU Wenyu;ZHANG Zhengpeng(School of Geomatics,Liaoning Technical University,Fuxin,Liaoning 123000,China)

机构地区：[1]辽宁工程技术大学测绘与地理科学学院,辽宁阜新123000

出　　处：《遥感信息》2020年第3期37-43,共7页Remote Sensing Information

摘　　要：针对遥感图像空间分辨率不足的问题,探讨了RAISR(rapid and accurate image super resolution)算法在单帧遥感图像超分辨率重建中的可行性。RAISR算法以处理自然图像为主,分为学习阶段和图像重建阶段。学习阶段是利用训练库图像生成滤波器,是算法的核心部分;图像重建阶段是利用滤波器重建图像。首先,在学习阶段,根据图像块的位置、角度、强度、相干性等特征对滤波器进行分类,并采用哈希列表存储;然后,针对遥感图像特点,优化了RAISR算法的滤波器尺寸,并采用USM(unsharp mask)方法增强边缘纹理特征,以达到最佳的重建效果;最后,用多组遥感图像进行了重建实验。结果表明:RAISR算法的重建质量与训练集图像的分辨率、数量、类别、所含地物类型有关;本文优化的RAISR算法重建后的遥感图像细节、边缘等信息都得到了改善。This paper discusses the feasibility of RAISR(rapid and accurate image super resolution)algorithm in single-frame remote sensing image super-resolution reconstruction to solve the problem that remote sensing image is affected by insufficient spatial resolution.RAISR focuses on processing natural images and is divided into a learning section and an image reconstruction section.The learning section mainly uses the training library image to generate filters,which is the core part of the algorithm,and the image reconstruction section uses filters to reconstruct the image.Firstly,in the learning section,the filter is classified according to the location,angle,intensity,coherence and other characteristics of the image block,and is stored in Hash list.Secondly,according to the characteristics of remote sensing image,the filter size is optimized,and USM(unsharp mask)method is used to enhance the edge texture features to achieve the best reconstruction effect.Finally,the reconstruction experiments are carried out with several groups of remote sensing images.The results show that:the reconstruction quality of the RAISR algorithm is related to resolution,quantity,category,and type of features of the training set images;the optimized RAISR algorithm in this paper can reconstruct the details and edges of the remote sensing image.

关 键 词：超分辨率重建 RAISR 滤波器 可行性 图像质量 

分 类 号：P236[天文地球—摄影测量与遥感]