| 基于小波变换理论的影像融合及评价 |
| |
| 引用本文： | 王坚,张继贤,刘正军,丁燕梅.基于小波变换理论的影像融合及评价[J].遥感信息,2006(6):I0002-I0003. |
| |
| 作者姓名： | 王坚 张继贤 刘正军 丁燕梅 |
| |
| 作者单位： | 1. 中国测绘科学研究院,摄影测量与遥感研究所,北京,100039;中国矿业大学,环境与测绘学院,徐州,221008
2. 中国测绘科学研究院,摄影测量与遥感研究所,北京,100039 |
| |
| 基金项目： | “科技部科研院所社会公益研究专项(2003DIA6N015、2004DE100625)” |
| |
| 摘 要： | 研究了基于MALLAT小波变换与静态小波变换(SWT)改进的IHS影像融合方案,并与标准IHS,BROVEY,PCA融合算法进行对比,分析了各种融合结果的优缺点。采用信息熵,标准差指标对信息量进行比较。选择平均值,相关系数,标准偏差指数进行融合影像光谱特征评价。针对ETM 全色波段与多光谱波段进行实验分析,结果表明小波变换融合算法总体优于传统融合算法,并能较好的保持光谱特性,但两种不同的离散小波变换方法改进的融合效果各有优缺点。
|
| 关 键 词： | MALLAT算法 静态小波 影像融合 |
| 文章编号： | 1000-3177（2006）88-0017-03 |
| 收稿时间： | 2005-10-31 |
| 修稿时间： | 2005-10-31 |
Image Merger Based on Wavelet Transform Theory and its Evaluation |
| |
| WANG Jian,ZHANG Ji-xian,LIU Zheng-jun,DING Yan-mei.Image Merger Based on Wavelet Transform Theory and its Evaluation[J].Remote Sensing Information,2006(6):I0002-I0003. |
| |
| Authors: | WANG Jian ZHANG Ji-xian LIU Zheng-jun DING Yan-mei |
| |
| Affiliation: | 1. Chinese Academy of Surveying and Mapping, Beijing , 100039 ; 2.College of Environment and Spatial Informatics, CUMT, Xuzhou, China , 221008 |
| |
| Abstract: | Improved IHS mergers based on Mallat Transform algorithm and Stationary Wavelet Transform(SWT) are discussed in this paper compared with standard IHS,Brovey,and PCA megers.Different mergers show different spatial resolution and spectral content of the resulting image.This work investigates the information entropy,standard variation,correlation coefficient,bias index and warping degree index for quantitative evaluating the fused images.Experimental results of distinct fusion methods are demonstrated using panchromatic and multi-spectral bands of ETM images.The wavelet transform based mergers are better than others as they have little spectral distortion.Mallat Transform based merger preserves more spatial information,but more spectral information is kept by Stationary Wavelet based merger. |
| |
| Keywords: | Mallat algorithm stationary wavelet image fusion |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
