一种基于HSI和小波变换的遥感图像融合方法  

A NEW REMOTE SENSING IMAGE FUSION METHOD BASED ON HSI AND WAVELET TRANSFORMATION

作　　者：刘建伟[1] 宋梦馨[2] 郭平[2] 

机构地区：[1]西安工业大学数理系,西安710032 [2]北京师范大学图形图像与模式识别实验室,北京100875

出　　处：《北京师范大学学报（自然科学版）》2008年第6期595-598,F0002,共5页Journal of Beijing Normal University(Natural Science)

基　　金：国家自然科学基金资助项目(60675011)

摘　　要：在对同一地区低分辨率的多光谱图像与高分辨率的全色图像的融合中,单独采用传统的HSI(hue-saturation-intensity)图像融合方法会出现颜色失真问题和传统的小波融合中存在分块模糊现象.为了改进融合效果,本文提出了一种新的融合方法,将HSI和小波变换进行有机结合,首先通过HSI变换获得多光谱图像的亮度成分,采用改进的基于主分量分析方法的融合规则将它与全色图像进行小波融合,得到新的亮度分量.将新的亮度分量进行增强处理后,进行HSI逆变换获得最终融合图像.实验结果证明,该方法不但提高了融合图像的空间分辨率,突出了目标的特征信息,而且更好地保留了原有多光谱图像的光谱性质.This paper proposes a new image fusion method for low-resolution multi-spectral image and high-resolution panchromatic image of the same area. Traditional HSI （ hue-saturation-intensity ） transformation method faces a problem namely color distortion, and traditional wavelet transformation has a problem of blurring. Therefore,this paper adopts a new method based on HSI and wavelet transform to solve the problem. First, the multi-spectral image is transformed into HSI space. Then, the intensity component of multi-spectral image and panchromatic image are merged by wavelet fusion, adopted a new fusion rule based on improved PCA （principal component analysis）. Finally, the fusion image is obtained by inverse HSI transformation. The experimental results show that the proposed method is better than traditional HSI method and wavelet transformation. It not only enhances spatial resolution of fusion image and obtains detail and feature information, but also preserves spectral information of the original multi-spectral image well.

关 键 词：遥感图像融合 HSI 小波变换 主分量分析 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置] TP391.41[自动化与计算机技术—控制科学与工程]