首页 | 本学科首页   官方微博 | 高级检索  
     

基于Curvelet多尺度几何分解的快速LHS多传感器图像融合?
引用本文:杨晓慧,李登峰.基于Curvelet多尺度几何分解的快速LHS多传感器图像融合?[J].传感技术学报,2009,22(4).
作者姓名:杨晓慧  李登峰
作者单位:河南大学数学与信息学院应用数学研究所,河南,开封,475004
基金项目:国家自然科学基金,河南省创新型科技人才队伍建设工程资助项目,河南省教育厅自然科学基金 
摘    要:为了解决IHS变换在多光谱和高分辨遥感图像融合中存在的光谱失真等问题,提出一种基于Curvelet多尺度几何分解的快速LHS融合算法.该算法采用快速LHS变换以更好地保持图像的色彩,并进一步改进运行速度;引入第二代Cur-velet多尺度几何分解则更有效地捕获图像中方向性的几何结构信息,并基于非平均分布间隔快速Fourier变换加快Curvelet的实现速度.实验结果表明:本文的图像融合算法在图像质量和运行速度上都获得了改进,融合图像具有较理想的空域和光谱分辨率,优于传统IHS变换、PCA,以及相应的小波变换结合快速LHS的融合结果.

关 键 词:多传感器  图像融合  快速LHS变换  Curvelet变换

Multisensor Image Fusion Using Fast LHS Based on Curvelet Mutiresolution Geometric Decomposition
YANG Xiao-hui,Li Deng-feng.Multisensor Image Fusion Using Fast LHS Based on Curvelet Mutiresolution Geometric Decomposition[J].Journal of Transduction Technology,2009,22(4).
Authors:YANG Xiao-hui  Li Deng-feng
Abstract:There is a problem of spectral distortion in multispectral and high-resolution image fusion based on IHS methods.To solve this problem, fast LHS transform is adopted to better hold the colours of source images and to improve the performance speed. On the other hand, the second generation Curvelet transform is introduced to capture the detail information effectively. Moreover, the digital transformation is based on unequally-spaced fast Fourier transform for speeding. Experimental results show that the proposed fusion method have improvements both in image quality and processing speed. The fused image is superior to that of based on IHS transform, PCA and wavelet transform combined with fast LHS transform. Our fusion results have ideal spatial and spectral resolution simultaneously both in visual and objective indexes.
Keywords:multi-sensors  image fusion  fast LHS transform  Curvelet transform
本文献已被 维普 万方数据 等数据库收录!
点击此处可从《传感技术学报》浏览原始摘要信息
点击此处可从《传感技术学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号