多角度高光谱遥感图像超分辨率重建研究

项目名称： 多角度高光谱遥感图像超分辨率重建研究

项目编号： No.61201342

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 电子学与信息系统

项目作者： 张洪艳

作者单位： 武汉大学

项目金额： 24万元

中文摘要： 高光谱遥感图像可以提供连续的光谱曲线，能够完整反映不同地物的诊断性光谱特征，在诸多领域具有广泛的应用，但较低的空间分辨率往往限制了其应用的范围与潜力。本项研究通过对同一平台相同类型传感器获取的多角度高光谱遥感图像进行超分辨率重建，将各幅图像之间的互补信息进行充分融合，有效提高高光谱图像的空间分辨率。针对多角度高光谱遥感图像超分辨率重建中存在的空间伪痕、光谱畸变等问题，发展稳健、快速的多角度高光谱遥感图像超分辨率重建方法，同时实现多角度高光谱图像空间信息的充分融合、光谱特性的有效保护和海量数据的快速处理。拟发展的方法能够在有效保护高光谱图像光谱特性的同时提高其空间分辨率，基于现有遥感观测系统提升图像的解译能力与实际应用潜力，在军事侦察、伪装识别、地质调查、精细农业等方面具有重要应用价值。

中文关键词： 多角度；高光谱；超分辨率重建；图像融合；

英文摘要： Hyperspectral remote sensing image has been widely used in many fields, thanks to its continuous spectral signature and pertinent spectral features, with which different land-covers can be effectively distinguished. However, the coarse spatial resolution of hyperspectral remote sensing image often limits the range and potential of its applications. This proposal intends to investigate the super-resolution reconstruction of multi-angular hyperpsectral image, in order to fully fuse the complementary information of different images acquired by similar sensors on the same platform with different angles and efficiently refine the spatial resolution of hyperspectral images. Considering the existing spatial artifacts and spectral distortion problems of multi-angular hyperspectral image super-resolution reconstruction, this proposal tries to develop a robust and fast multi-angular hyperspectral image super-resolution reconstruction method and simultaneously realize the fusion of spatial information, preservation of spectral properties and fast processing of massive data. This proposal aims at efficiently improving the spatial resolution and preserving the spectral properties of hyperspectral image, as well as enhancing its interpretation capabilities and practical application potentials with the existing remote sensing

英文关键词： Multi-angle；Hyperspectral；Super-resolution reconstruction；Image fusion；