基于PCA+àstrous小波融合方法  

Remote sensing image fusion based on PCA+àtrous wavelet transform

作　　者：罗艳[1] 刘荣[1] 朱煜峰[1] 

机构地区：[1]东华理工大学地测学院,江西抚州344000

出　　处：《测绘科学》2009年第6期108-109,290,共3页Science of Surveying and Mapping

基　　金：国家自然基金(40401050);核资源与环境教育部重点实验室(070704东华理工大学);江西省数字国土实验室(DLLJ200804)资助

摘　　要：针对PCA变换融合影像存在较严重的光谱失真现象,以及àstrous小波融合影像保真度高,而空间分辨率相对低的情况,本文提出一种基于PCA+àstrous小波融合算法。新方法首先对将多波段图像经PCA变换至各不相关的成分,而后对高分辨率图像与低分辨率图像主成分按照特定融合规则进行融合处理,并使用该融合后的第一主成份分量来替代高分辨率图像与低分辨率图像进行àstrous小波融合,即PCA变换与àstrous小波变换相结合的融合处理方法。主观视觉分析和客观参数表明,新方法不仅很好的保留了影像的光谱信息,而且兼顾了地物细节能力的表达。For the spectral distortion of PCA transform and low spatial resolution of atrous wavelet transform, a new technique, PCA ＋ astrous based wavelet fusion algorithm was developed in this paper. Firstly a fusion image using PCA transform to merge multispectral image and high-resolution panchromatic image was obtained. Then new principle components for the new muhispecral image were merged. Using htrous wavelet merger and PCA merger, the fusion of muhiresolution image and the first component of the fused image were performed. The new method is capable of preserving its spectral content while enhancing the spatial quality of the multispectral image to a greater extent.

关 键 词：影像融合 高空间分辨率 小波变换 评定准则 PCA 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]