光子学报 ›› 2020, Vol. 49 ›› Issue (10): 1010001-1010001.doi: 10.3788/gzxb20204910.1010001

• 图像处理 • 上一篇    下一篇

结合光谱响应函数和全局方差匹配的遥感图像融合

李俊杰(), 傅俏燕, 姜涛   

  1. 中国资源卫星应用中心,北京 100094
  • 收稿日期:2020-06-30 接受日期:2020-08-19 出版日期:2020-10-25 发布日期:2020-10-13
  • 作者简介:李俊杰(1983—),男,高级工程师,硕士,主要研究方向为遥感数据处理、图像融合. Email: lijunjie299@126.com
  • 基金资助:
    国家重点研发计划(2018YFB0505000)

Remote Sensing Image Fusion Based on Spectral Response Function and Global Variance Matching

Jun-jie LI(), Qiao-yang FU, Tao JIANG   

  1. China Centre for Resources Satellite Data and Application,Beijing 100094,China
  • Received:2020-06-30 Accepted:2020-08-19 Online:2020-10-25 Published:2020-10-13
  • Supported by:
    National Key Research and Development Program of China(2018YFB0505000)

摘要:

为保持空间细节和减少光谱扭曲,同时针对改进的分量替换融合方法在构建强度分量时,容易存在系数为负或过小的问题,提出了一种结合光谱响应函数和全局方差匹配的遥感图像融合方法.该方法基于通用分量替换融合框架,使用全色和多光谱传感器的光谱响应函数反映的辐射能量响应的比例关系来构造强度分量,物理意义明确,数学形式简单明了.同时使用全局协方差和方差之比来计算空间细节调制参数,减少光谱畸变,满足通用分量替换融合框架的约束条件.选取两组不同卫星图像作为测试数据,并与多种成熟的融合方法相对比,结果表明,该方法得到的融合图像空间和光谱质量都较好.

关键词: 遥感图像融合, 分量替换融合, 光谱响应函数, 全局方差匹配, 多光谱图像

Abstract:

In order to keep the spatial details and reduce the spectral distortion, and to solve the problem that the coefficients of the improved component replacement fusion method are often negative or too small when building the intensity components, a remote sensing image fusion method combining spectral response function and global variance matching is proposed. Based on the general component replacement fusion framework, the intensity component is constructed by using the proportional relationship of the radiation energy response reflected by the spectral response function of panchromatic and multispectral sensors. The physical meaning is explicit, and the mathematical form is simple and clear. At the same time, the spatial detail modulation parameters are determined by using the ratio of global covariance to variance to reduce the spectral distortion and meet the constraints of the general component replacement fusion framework. The proposed method is compared with many mature fusion methods on two groups of different satellite image data, the results show that the fusion image spatial and spectral quality are better.

Key words: Remote sensing image fusion, Component substitution pansharpening, Spectral response function, Global variance matching, Multispectral image

中图分类号: