项目名称: 基于同质区分析的高光谱影像混合像元稀疏分解研究
项目编号: No.61501200
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 无线电电子学、电信技术
项目作者: 孔祥兵
作者单位: 黄河水利委员会黄河水利科学研究院
项目金额: 22万元
中文摘要: 由于地物分布的复杂性和传感器空间分辨率的有限性,混合像元普遍存在于高光谱影像中,是高光谱影像定量分析和应用的最大障碍。高光谱影像含有丰富的空间、光谱和辐射等信息,然而常见的混合像元分解方法多是将高光谱影像视为离散而无任何关系的一个个高维像元光谱集,未能充分挖掘并高效利用像元间的空间相关性;且混合像元分解过程和结果往往不能准确反映影像空间地物分布特征和客观实际。针对上述问题,本项目从像元光谱相似性分析视角出发,以高光谱影像同质区分析为基础,以构建新型的高光谱影像端元光谱提取和混合像元约束分解模型为目的,借鉴先进的数学理论和影像分析思想,初步提出并完善结合影像空间信息和像元光谱信息的高光谱影像混合像元稀疏分解的相关支撑理论和关键技术。本项目的研究成果将会进一步促进高光谱遥感影像走向定量分析和应用,对我国的资源普查、环境与灾害高效监测和深空探测也将具有十分重要的现实意义。
中文关键词: 同质区分析;混合像元分解;空间信息;端元提取;非负矩阵分解
英文摘要: There are lots of mixed pixels in hyperspectral imagery, which is the biggest obstacle of quantitative analysis and application of hyperspectral data, due to the distribution complexity of the materials and the low spatial resolution of the sensors. However, most approaches have been designed from a spectroscopic viewpoint and thus, tend to neglect the existing spatial correlation between pixels,and the hyperspectral sparse unmixing problem need further reserch conserned the distribution of spatial object and the objective reality. In response to above issues, the project aims to present novel models of hyperspectral sparse unmixing based on image spatial information and spectral information characteristics analysis, and to improve related theories and key technologies with novel mathematical theory and image analysis ideas. The research topics will further promote quantitative analysis and application of hyperspectral imagery, and will also have a very important practical significance for our resources survey,environmental disasters efficient monitoring and deep space exploration.
英文关键词: homogeneous-region analysis;mixed pixel unmixing;spatial information;enndmember extraction;nonnegative matrix factorization