融合SPOT5与Landsat的超分辨率技术：提升空间分辨率

5星 · 超过95%的资源 150 浏览量更新于2024-08-30 1 收藏 1.77MB PDF 举报

"通过基于学习的超分辨率，通过与SPOT5图像融合来提高Landsat TM / ETM的空间分辨率" 本文主要探讨了一种利用学习方法来融合Landsat Thematic Mapper (TM) 和 Enhanced Thematic Mapper Plus (ETM+) 图像与Systeme Pour l'Observation de la Terre 5 (SPOT5) 图像，以提升Landsat图像的空间分辨率的技术。Landsat TM/ETM+图像具有宽幅宽度的优势，而SPOT5图像则以其高空间分辨率著称。通过融合这两类图像，可以结合两者的优点，生成具有宽幅和高分辨率特征的融合图像。作者首先建立了图像降级模型，通过模糊和下采样操作模拟从SPOT5图像到Landsat TM/ETM+图像的成像过程。这个模型允许他们从每个SPOT5图像生成一个模拟的Landsat图像，这一步骤消除了对输入图像几何配准的需求，简化了处理流程。接下来，融合过程分为两个阶段。在第一阶段，他们学习到一个字典对，该字典对能代表高分辨率和低分辨率图像的细节。这个字典对是通过对给定的SPOT5图像和模拟的TM/ETM+图像进行学习得到的。在第二阶段，他们应用字典对和稀疏编码算法，对输入的Landsat图像进行超分辨率处理，从而提高其空间分辨率。值得注意的是，这种方法不仅可以用于融合任务，还能处理TM/ETM+和SPOT5图像的常规空间和频谱融合问题。这项工作展示了如何通过学习和超分辨率技术有效地结合不同卫星图像的数据特性，以提高遥感图像的分析能力。对于环境监测、土地覆盖变化分析等应用，这种融合技术可以提供更精确、更详细的信息。通过这种方式，Landsat TM/ETM+图像的条带宽度优势和SPOT5图像的高分辨率特性可以被充分利用，为地球观测和地理信息系统等领域带来显著的改进。在实际应用中，这种融合技术可能需要针对不同的地理区域和应用场景进行调整，以优化结果。同时，它也可能对图像处理软件和计算资源有较高的要求，因为超分辨率过程通常涉及复杂的数学运算。尽管如此，这项研究为提高遥感图像处理效率和精度开辟了新的途径，对于未来遥感技术的发展具有重要的理论和实践意义。

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 53, NO. 3, MARCH 2015 1195

Improving the Spatial Resolution of Landsat

TM/ETM+ Through Fusion With SPOT5

Images via Learning-Based Super-Resolution

Huihui Song, Bo Huang, Member, IEEE, Qingshan Liu, and Kaihua Zhang

Abstract—To take advantage of the wide swath width of Land-

sat Thematic Mapper (TM)/Enhanced Thematic Mapper Plus

(ETM+) images and the high spatial resolution of Système Pour

l’Observation de la Terre 5 (SPOT5) images, we present a learning-

based super-resolution method to fuse these two data types. The

fused images are expected to be characterized by the swath width

of TM/ETM+ images and the spatial resolution of SPOT5 images.

To this end, we ﬁrst model the imaging process from a SPOT image

to a TM/ETM+ image at their corresponding bands, by building

an image degradation model via blurring and downsampling oper-

ations. With this degradation model, we can generate a simulated

Landsat image from each SPOT5 image, thereby avoiding the

requirement for geometric coregistration for the two input images.

Then, band by band, image fusion can be implemented in two

stages: 1) learning a dictionary pair representing the high- and

low-resolution details from the given SPOT5 and the simulated

TM/ETM+ images; 2) super-resolving the input Landsat images

based on the dictionary pair and a sparse coding algorithm. It

is noteworthy that the proposed method can also deal with the

conventional spatial and spectral fusion of TM/ETM+ and SPOT5

images by using the learned dictionary pairs. To examine the

performance of the proposed method of fusing the swath width

of TM/ETM+ and the spatial resolution of SPOT5, we illustrate

the fusion results on the actual TM images and compare with

several classic pansharpening methods by assuming that the cor-

responding SPOT5 panchromatic image exists. Furthermore, we

implement the classiﬁcation experiments on both actual images

and fusion results to demonstrate the beneﬁts of the proposed

method for further classiﬁcation applications.

Index Terms—Landsat Thematic Mapper (TM) or Enhanced

Thematic Mapper Plus (ETM+) image, spatial resolution,

Système Pour l’Observation de la Terre 5 (SPOT5) image,

super-resolution, swath width.

I. INTRODUCTION

ITH the emergence of a wide variety of remote sensing

instruments and the widespread application of remote

Manuscript received September 9, 2013; revised January 16, 2014 and

April 17, 2014; accepted June 3, 2014. The work of H. Song and B. Huang

was supported by the Hong Kong Research Grant Council under Grant CUHK

444612. The work of Q. Liu was supported in part by the National Natural

Science Foundation of China under Grant 61272223 and in part by the National

Science Foundation of Jiangsu Province under Grant BK2012045. The work of

K. Zhang was supported by the Nanjing University of Information Science and

Technology Scientiﬁc Research Foundation under Grant S8113049001.

H. Song, Q. Liu, and K. Zhang are with the Nanjing University of Informa-

tion Science and Technology, Nanjing 210044, China (e-mail: hhsongsherry@

gmail.com).

B. Huang is with the Chinese University of Hong Kong, Shatin, Hong Kong.

Color versions of one or more of the ﬁgures in this paper are available online

at http://ieeexplore.ieee.org.

Digital Object Identiﬁer 10.1109/TGRS.2014.2335818

sensing data, multisensor multiresolution image fusion is be-

coming increasingly more important [1]. A common feature

among these satellite sensors is that they trade off between

spatial resolution and other sensor properties, including spectral

resolution, temporal resolution, and swath width, caused by

technical and budget limitations. Two examples are the The-

matic Mapper (TM) or the Enhanced Thematic Mapper Plus

(ETM+) sensor aboard the Landsat satellite (we will take the

Landsat TM sensor as an example in this paper) and the High

Resolution Geometrical (HRG) instrument on the commercial

satellite Système Pour l’Observation de la Terre 5 (SPOT5).

Although launched by different agencies, SPOT (1–6 have

currently been launched) and Landsat (1–8 have currently been

launched) satellites have provided large amounts of data for

various ﬁelds since the launch of the ﬁrst one. With the launch

of the new SPOT and Landsat satellites i n the future, studying

the fusion of data from these two platforms will greatly promote

their applications.

Landsat TM provides unique resources for global change

research and applications in such areas as agriculture, cartog-

raphy, and geology [2], with a spatial resolution of 30 m, at

most, of the seven bands (except band 6 with a spatial resolution

of 120 m) and a swath width of 185 km, whereas SPOT5

HRG aims for high-spatial-resolution images featured by four

multispectral bands [green, red, near-infrared (NIR) with a

spatial resolution of 10 m, and short-wave infrared (SWIR)

with a spatial resolution of 20 m] and one panchromatic band

with a swath width of 60 km. Obviously, these two sensors

are complementary in their spatial resolution and swath width

properties, i.e., the TM sensor is characterized by lower spatial

resolution but wider swath width, whereas the SPOT5 sensor

is characterized by higher spatial resolution but narrower swath

width. However, higher spatial resolution and a greater cover

area may be simultaneously needed in practical applications,

e.g., in large-area land cover/land use mapping with high spatial

resolution [3], [4] or in cases where higher spatial resolution

data are required but only low-resolution TM/ETM+ data are

available for the study scene. Thus, we propose to enhance the

spatial resolution of TM images by merging with an available

SPOT5 image.

In fact, the fusion of Landsat TM/ETM+ and SPOT5 images

has been explored by many researchers in the past two decades

[5]–[7]. However, most of them focused mainly on the fusion

of TM/ETM+ multispectral and SPOT5 panchromatic images

to integrate the good spectral property of the former and the

good spatial property of the latter, respectively. This spatial and

IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.

See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

下载后可阅读完整内容，剩余9页未读，立即下载

weixin_38636577

粉丝: 4
资源: 935

融合SPOT5与Landsat的超分辨率技术：提升空间分辨率

Landsat 8 OLI多光谱与全色影像融合算法的比较*

Landsat 5辐射定标并波段合成

mosaic方法实现图像融合

LandSat TM/ETM论文

遥感图像处理：传感器.pdf

遥感原理与方法课程实践教学改革探索.docx

遥感考试样卷与答案解析（详细整理版）.doc

遥感原理与应用 课件 武汉大学 第二章 遥感平台及运行特点

全球遥感卫星与传感器技术概览

遥感技术基础：复习要点与关键词

最新资源

遥感原理与应用课件武汉大学第二章遥感平台及运行特点