摘要混合像元分解是高光谱遥感应用技术中公认的难题,主要是由于传感器的空间分辨率有限,同时自然界中的地貌是复杂多样的 ,尤其是地物本身的尺寸就比较小,导致像元很少由单一地表覆盖类型组成,通常都是几种地物的混合。所以解决混合像元分解的问题是必须的,如此才可以提高遥感应用的精度。27284
本文通过对高光谱遥感图像混合光谱的分解问题进行了总结,探索了混合光谱产生的主要原因,并对其中的端元提取进行了深入研究,分析了端元提取的原理及研究现状,以及云计算和Hadoop技术用于高光谱遥感数据处理的潜力。在此基础上对现有的PPI(纯净像元指数)端元提取算法进行了研究,并提出了基于Hadoop云计算平台进行端元提取并行优化的思路,设计了一个高光谱图像端元提取系统,通过优化设计PPI高光谱图像端元提取算法,提升了算法的处理效率,同时实现了Hadoop进程的有效控制,提供简易的操作界面以及清晰的数据展示。
关键词 高光谱遥感, 混合像元分解,Hadoop,云计算,PPI
毕业论文设计说明书外文摘要
Title The Hyperspectral Image Endmember Extraction Algorithm Based on Hadoop and Cloud Computing
Abstract
Unmixing is always a difficuty of the Hyperspectral Remote Sensing applications,because the spatial resolution of sensor is limited,and the nature is complex,inparticular,the size of the feature itself is always small,making that pixel is hardly made up of single land cover types,it is usually a mix of some feature.So we must solve the problem of unmixing,to improve the accuracy of remote sensing applications.
Based on the decomposition problem of Hyperspectral remote sensing image MixingSpectral,this article explores the main reason for the MixingSpectral producted,and conducted in-depth research for the Endmember extraction,analyzes the pinciple and research development of Endmember extraction,as well as the potential of cloud computing and Hadoop technology for hyperspectral remote sensing data processing.On this basis,studied for the existing PPI(pixel pure index) Endmember extraction algorithm,and proposed an idea that perform endmember extraction parallel optimization based on Hadoop and cloud platform,designed a system of hyperspectral image endmember extraction, By optimizing and designing PPI hyperspectral image endmember extraction algorithm, enhanced the processing efficiency of the algorithm, while achieving the effective control of Hadoop process, and providing a simple user interface and clearly data showing.
Keywords:Hyperspectral remote sensing,Unmixing,Hadoop,Cloud computing,PPI
目 次
1 绪论 1
1.1 研究背景 1
1.2 国内外研究现状 1
1.3 研究目的和意义 2
1.4 论文研究内容 3
2 相关概念与技术 5
2.1 高光谱遥感简介 5
2.2 云计算简介 5
2.3 Hadoop简介 7
2.4 端元提取简介 8
3 基于HADOOP云计算的高光谱图像端元提取系统需求分析与系统设计 10
3.1 可行性分析 10
3.2 需求分析 10
3.3 PPI端元提取算法的分布式并行优化设计 11
3.4 管理系统设计 12
4 基于HADOOP云计算的高光谱图像端元提取系统具体实现 14
4.1 Hadoop平台搭建的实现 14
4.2 端元提取算法的优化与实现 20 基于Hadoop云计算的高光谱图像端元提取算法设计与实现:http://www.youerw.com/jisuanji/lunwen_21745.html