基于Curvelet和改进区域方差的遥感图像融合  

Remote sensing images fusion based on curvelet transform and improved regional variance weighting

作　　者：栾静[1] 殷明[1] 于立萍[1] 白瑞峰[1] 水珺 

机构地区：[1]合肥工业大学数学学院,安徽合肥230009

出　　处：《合肥工业大学学报（自然科学版）》2015年第9期1220-1225,共6页Journal of Hefei University of Technology：Natural Science

基　　金：国家自然科学基金资助项目(11172086);安徽省自然科学基金资助项目(1308085MA09);安徽省教育厅自然科学研究重点资助项目(KJ2013A216)

摘　　要：为了同时提高多光谱图像与全色图像融合的空间分辨率和保留较好的图像光谱性,文章提出了一种基于Curvelet变换的图像融合新方法。首先对多光谱图像进行HIS变换,获得亮度I分量,对全色图像和亮度分量I进行直方图匹配,再对I分量进行平滑滤波;然后对全色图像与I分量进行Curvelet分解,分别得到低频与高频系数,对低频子带系数的融合规则采用改进的区域方差加权分析法,对高频子带系数采用拉普拉斯能量和算子结合改进的加权平均的融合规则;最后进行Curvelet逆变换和HIS逆变换得到融合图像。仿真实验得到了较好的客观评价数据和视觉效果。In order to simultaneously improve the spatial resolution of multispectral image and panchro‐matic image fusion and keep good image spectrum ,a new image fusion method based on curvelet transform is proposed .Firstly ,HIS transform is applied in multispectral image for getting intensity component I ,histogram matching for panchromatic image and intensity component I is done ,and smoothing filtering of the intensity component I is also done .Secondly ,the curvelet decomposition of the panchromatic image and intensity component I is conducted for getting low frequency and high fre‐quency coefficients respectively ,the improved regional variance weighting analysis method is used for the fusion rule of the low frequency sub‐band coefficients ,and the sum‐modified Laplacian combined with improved weighted average for the high frequency sub‐band coefficients .Finally ,the fused image is obtained by curvelet inverse transform and HIS inverse transform .The simulation results show the better objective evaluation data and visual effect .

关 键 词：遥感图像 HIS变换 CURVELET变换 拉普拉斯能量和 改进的区域方差加权 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]