IKONOS影像的最佳融合技术研究  

Study on optimum fusion technology of IKONOS image

作　　者：宋刚贤[1] 潘剑君[2] 朱文娟[2] 

机构地区：[1]南京农业大学公共管理学院,南京210095 [2]南京农业大学资源与环境学院,南京210095

出　　处：《测绘科学》2009年第2期106-108,共3页Science of Surveying and Mapping

摘　　要：在研究遥感数字影像融合算法和原理的基础上,选择高空间分辨率的IKONOS遥感影像作为数据源,采用能够以全波段融合的主成分变换、乘积变换、Gram-schmidt变换和小波变换四种融合方法进行比较研究,并通过主观和客观两方面对融合效果进行评价。结果表明,Gram-schmidt变换法融合的IKONOS影像光谱质量最好,同时也较高程度地保持了高几何分辨率的空间信息,是权衡光谱信息和空间信息综合效应最佳的一种方法。Many applications in remote sensing require fusion of low-resolution multi-spectral of hyper-spectral images with highresolution panchromatic images to identify map materials at high resolution. A number of methods are currently used to merge images. This paper compared the performances of four fusion approaches （PCA, Product, Gram-schmidt and wavelet transformation） to each other by means of qualitative and quantitative analysis. The results showed that Gram-sehmidt was the optimum approach among the four techniques for it preserves the highest spectral quality and spatial resolution information.

关 键 词：遥感影像 IKONOS 融合 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]