利用小波变换对影像进行融合的研究

王智均; 李德仁; 李清泉

美国《工程索引》（EI）核心期刊
美国《剑桥科学文摘》（CSA）收录期刊
荷兰斯高帕斯数据库（Scopus）收录期刊
英国INSPEC数据库（SA）收录期刊

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

利用小波变换对影像进行融合的研究

王智均, 李德仁, 李清泉

文章导航 > 武汉大学学报 ● 信息科学版 > 2000 > 25(2): 137-142

王智均, 李德仁, 李清泉. 利用小波变换对影像进行融合的研究[J]. 武汉大学学报 ● 信息科学版, 2000, 25(2): 137-142.

引用本文:

王智均, 李德仁, 李清泉. 利用小波变换对影像进行融合的研究[J]. 武汉大学学报 ● 信息科学版, 2000, 25(2): 137-142.

WANG Zhijun, LI Deren, LI Qingquan. Image Fusion with Wavelets Analysis[J]. Geomatics and Information Science of Wuhan University, 2000, 25(2): 137-142.

Citation:

WANG Zhijun, LI Deren, LI Qingquan. Image Fusion with Wavelets Analysis[J]. Geomatics and Information Science of Wuhan University, 2000, 25(2): 137-142.

利用小波变换对影像进行融合的研究

1 武汉测绘科技大学测绘遥感信息工程国家重点实验室, 武汉市珞喻路129号, 430079;
2 武汉测绘科技大学校长办公室, 武汉市珞喻路129号, 430079

基金项目: 国家自然科学基金,编号49631050

详细信息

作者简介:
王智均,男,31岁,博士生,现从事遥感图像处理、GIS研究。E-mail:zjwang@wtusm.edu.cn

中图分类号: TP751;P237.3

Image Fusion with Wavelets Analysis

1 National Laboratory for Information Engineering in Surveying, Mapping and Remote Sensing, WTUSM, 129 Luoyu Road, Wuhan, China, 430079;
2 Presidential Secretariat, WTUSM, 129 Luoyu Road, Wuhan, China, 430079;

摘要: 基于小波变换可以对影像进行正交分解,而不丢失原来信号所包含的信息。提出了一种Wallis变换、小波变换和IHS变换相结合用于融合的方法,该方法可以有效地提高多光谱影像的空间分辨率,同时保持原来多光谱影像的色调。
- 影像融合 /
- 小波变换 /
- 多分辨率分析 /
- Mallat算法 /
- 频率域 /
- Wallis变换
Abstract: The definition of image fusion is described by a figure. The evolution of image fusion technique is reviewed and the adventage and shortcoming of those image fusion methods are briefly explained as well. By analyzing formulation theory of remote sensing satellite image in frequency domain, it is concluded that high-resolution image covers more wide frequency domain and low-resolution multi-spectral image covers relatively smaller frequency domain. The requirement for combining the high frequency part of high resolution image with the low frequency part of the low resolution multi-spectral image is arisen from several user groups. Some possibilities to combine the high frequency part of the high resolution images with the low resolution images which are called image fusion methods are proposed by different application purposes. The theoretical basis of image fusion is then given. The basic theory and advantage of wavelets, Mallat algorithm and Wallis transformation are introduced. Wavelets analysis can be used to decompose the image orthogonal while keeping the information of the original image and Wallis transformation can be used to enhance the contrast of weak contrast area and remove the noise. It is found that those benefits of wavelet analysis, Wallis transformation can be used for image fusion. IHS transformation is a common image fusion technique. It can be used to transfer RGB image channels from RGB space to IHS channels of IHS space, the intensity of the image is separated from hue and saturation channels. It provides the possibility to keep the colorful characteristics of the original color composite while modulating the intensity channel. An image fusion method which integrates the advantages of Wallis transformation,wavelets analysis and IHS transformation is proposed. Median filter is also used to remove the noise of the hue and saturation channel. A test area of the suburban area of western Beijing city is chosen to check the proposed image fusion process. The test site includes different topographic geomorphies such as mountain area,rural area, lakes and urban area. It is useful to check the proposed image fusion method whether it can enhance the resolution by checking the details of the mountain area and the roads of urban area and keep the original color composite by checking whether the color of lake surface and the land cover is changed or not. Some quantitative and qualitative evaluation methods are introduced and the results show that details of high resolution image are added to the low resolution multi-spectral images while keeping the spectral characteristics of the original image.
- image fusion /
- wavelet transformation /
- multi-resolution analysis /
- Mallat algorithm /
- frequency domain /
- Wallis transformation

[1]	王达, 卞红雨. 水下声图像多分辨率融合研究 . 武汉大学学报 ● 信息科学版, 2015, 40(1): 77-82.
[2]	陈江平, 韩青, 胡晶, 张鹏林. 顾及小波变换的土地利用变化与经济因子的多尺度相关性分析 . 武汉大学学报 ● 信息科学版, 2013, 38(9): 1118-1121.
[3]	王伟, 宁津生, 罗志才, 汪海洪. 利用小波变换分离俯冲带重力异常 . 武汉大学学报 ● 信息科学版, 2012, 37(2): 187-190.
[4]	陈亮, 朱长青, 任娜, 朱一姝. 利用小波变换的视频GIS数据数字水印算法 . 武汉大学学报 ● 信息科学版, 2012, 37(10): 1256-1259.
[5]	黄登山, 杨敏华, 胥海威, 姚学恒. 利用最优估计理论进行多光谱与全色影像融合 . 武汉大学学报 ● 信息科学版, 2011, 36(9): 1039-1042.
[6]	宁津生, 王伟, 汪海洪, 罗佳. 应用小波变换确定琉球俯冲带的深部特征 . 武汉大学学报 ● 信息科学版, 2010, 35(10): 1135-1137.
[7]	林立宇, 茅力非, 王雷光, 秦前清. 一种基于分段线性系数变换的编码预处理算法 . 武汉大学学报 ● 信息科学版, 2008, 33(7): 680-683.
[8]	蔡昌盛, 高井祥. GPS周跳探测及修复的小波变换法 . 武汉大学学报 ● 信息科学版, 2007, 32(1): 39-42.
[9]	陈仁喜, 赵忠明, 王殷行. 基于整型小波变换的DEM数据压缩 . 武汉大学学报 ● 信息科学版, 2006, 31(4): 344-347.
[10]	黄昕, 张良培, 李平湘. 基于小波的高分辨率遥感影像纹理分类方法研究 . 武汉大学学报 ● 信息科学版, 2006, 31(1): 66-69.
[11]	邹勤, 贾永红. 一种基于形态金字塔的遥感影像融合方法及其性能评价 . 武汉大学学报 ● 信息科学版, 2006, 31(11): 971-974.
[12]	邵海梅, 梅天灿, 秦前清. 基于小波变换的扩频水印算法 . 武汉大学学报 ● 信息科学版, 2003, 28(5): 626-629.
[13]	李平湘, 王智均. 高空间分辨率影像和多光谱影像融合的研究 . 武汉大学学报 ● 信息科学版, 2003, 28(1): 14-17.
[14]	高戈, 胡瑞敏, 李德仁. 改进的基于小波变换的语音活动检测算法 . 武汉大学学报 ● 信息科学版, 2002, 27(1): 85-88.
[15]	王智均, 李德仁, 李清泉. Wallis变换在小波影像融合中的应用 . 武汉大学学报 ● 信息科学版, 2000, 25(4): 338-342.
[16]	耿则勋. 基于小波变换的遥感影像保持量测精度的压缩技术研究 . 武汉大学学报 ● 信息科学版, 1998, 23(2): 187-187.
[17]	仇彤. 基于小波变换的松弛法影像匹配 . 武汉大学学报 ● 信息科学版, 1998, 23(2): 145-148.
[18]	曾湧, 廖明生, 张剑清, 沈未名. 保持匹配一致性的遥感影像压缩编码 . 武汉大学学报 ● 信息科学版, 1997, 22(3): 245-247,251.
[19]	李军, 林宗坚. 基于二进小波变换的遥感影像镶嵌方法 . 武汉大学学报 ● 信息科学版, 1995, 20(4): 305-309,347.
[20]	邵巨良, 李德仁. 小波变换及其在影像表示中的应用和其方向选择性算法的改进 . 武汉大学学报 ● 信息科学版, 1993, 18(2): 1-9.

点击查看大图

计量

文章访问数: 574
HTML全文浏览量: 28
PDF下载量: 247
被引次数: 0

利用小波变换对影像进行融合的研究

1 武汉测绘科技大学测绘遥感信息工程国家重点实验室, 武汉市珞喻路129号, 430079;

2 武汉测绘科技大学校长办公室, 武汉市珞喻路129号, 430079

基金项目: 国家自然科学基金,编号49631050

作者简介:
王智均,男,31岁,博士生,现从事遥感图像处理、GIS研究。E-mail:zjwang@wtusm.edu.cn

收稿日期: 1999-09-14
刊出日期: 2000-02-05

中图分类号: TP751;P237.3

关键词:

频率域 /

摘要: 基于小波变换可以对影像进行正交分解,而不丢失原来信号所包含的信息。提出了一种Wallis变换、小波变换和IHS变换相结合用于融合的方法,该方法可以有效地提高多光谱影像的空间分辨率,同时保持原来多光谱影像的色调。

English Abstract

Image Fusion with Wavelets Analysis

1 National Laboratory for Information Engineering in Surveying, Mapping and Remote Sensing, WTUSM, 129 Luoyu Road, Wuhan, China, 430079;

2 Presidential Secretariat, WTUSM, 129 Luoyu Road, Wuhan, China, 430079;

Received Date: 1999-09-14
Publish Date: 2000-02-05

Keywords:

Abstract: The definition of image fusion is described by a figure. The evolution of image fusion technique is reviewed and the adventage and shortcoming of those image fusion methods are briefly explained as well. By analyzing formulation theory of remote sensing satellite image in frequency domain, it is concluded that high-resolution image covers more wide frequency domain and low-resolution multi-spectral image covers relatively smaller frequency domain. The requirement for combining the high frequency part of high resolution image with the low frequency part of the low resolution multi-spectral image is arisen from several user groups. Some possibilities to combine the high frequency part of the high resolution images with the low resolution images which are called image fusion methods are proposed by different application purposes. The theoretical basis of image fusion is then given. The basic theory and advantage of wavelets, Mallat algorithm and Wallis transformation are introduced. Wavelets analysis can be used to decompose the image orthogonal while keeping the information of the original image and Wallis transformation can be used to enhance the contrast of weak contrast area and remove the noise. It is found that those benefits of wavelet analysis, Wallis transformation can be used for image fusion. IHS transformation is a common image fusion technique. It can be used to transfer RGB image channels from RGB space to IHS channels of IHS space, the intensity of the image is separated from hue and saturation channels. It provides the possibility to keep the colorful characteristics of the original color composite while modulating the intensity channel. An image fusion method which integrates the advantages of Wallis transformation,wavelets analysis and IHS transformation is proposed. Median filter is also used to remove the noise of the hue and saturation channel. A test area of the suburban area of western Beijing city is chosen to check the proposed image fusion process. The test site includes different topographic geomorphies such as mountain area,rural area, lakes and urban area. It is useful to check the proposed image fusion method whether it can enhance the resolution by checking the details of the mountain area and the roads of urban area and keep the original color composite by checking whether the color of lake surface and the land cover is changed or not. Some quantitative and qualitative evaluation methods are introduced and the results show that details of high resolution image are added to the low resolution multi-spectral images while keeping the spectral characteristics of the original image.

王智均, 李德仁, 李清泉. 利用小波变换对影像进行融合的研究[J]. 武汉大学学报 ● 信息科学版, 2000, 25(2): 137-142.

引用本文:

王智均, 李德仁, 李清泉. 利用小波变换对影像进行融合的研究[J]. 武汉大学学报 ● 信息科学版, 2000, 25(2): 137-142.

WANG Zhijun, LI Deren, LI Qingquan. Image Fusion with Wavelets Analysis[J]. Geomatics and Information Science of Wuhan University, 2000, 25(2): 137-142.

Citation:

WANG Zhijun, LI Deren, LI Qingquan. Image Fusion with Wavelets Analysis[J]. Geomatics and Information Science of Wuhan University, 2000, 25(2): 137-142.