项目名称: 异源星载光学、SAR遥感影像的空间信息提取关键技术研究
项目编号: No.61301278
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 无线电电子学、电信技术
项目作者: 吴颖丹
作者单位: 湖北工业大学
项目金额: 25万元
中文摘要: 目前,针对异源星载光学、SAR遥感影像的空间信息提取,仍难以满足实际需求。关键原因在于两种影像的成像波谱、机理、方式完全不同,这造成影像精确定向定位和同名点匹配更加复杂和困难。针对该问题,本课题拟对其中的区域网平差、影像匹配等关键问题进行深入研究,具体内容包括:1)遥感影像有理函数模型参数的稳健无偏估计;2)基于改进误差补偿模型的异源影像有理函数模型区域网平差;3)顾及几何约束条件和高阶结构特征的异源遥感影像同名点自动匹配。对于前两项,通过对有理函数模型参数求解的优化和SAR影像误差补偿模型的研究,来实现异源遥感影像的精确几何定位;对于后者,通过有理函数区域网平差模型和超图匹配模型引入约束条件,实现同名点的精确可靠匹配。本研究有利于异源遥感影像在空间信息获取方面的优势互补,最大程度挖掘影像信息提取能力,具有较高的理论价值和应用价值。
中文关键词: 异源遥感影像;空间信息提取;有理函数模型;区域网平差;影像匹配
英文摘要: For the multi-sensor images from the spaceborne optical and SAR sensors, the spatial information extraction is still difficult to meet the actual demand. The main reason is that these two kinds of images are totally different in the imaging spectrum, mechanism and mode, which cause the block adjustment and image matching more complex and difficult. To solve the problem, the project intends to carry out the in-depth study on the key issues of bundle adjustment and corresponding points matching in the aerotriangulation, and they specifically includes: firstly, the robust and unbiased estimation of the Rational Function Model (RFM) parameters for remote sensing imagery. Secondly, RFM bundle adjustment based on the improved error compensation model for multi-sensor remote sensing images. And thirdly, robust image match incorporating the geometric restraints and high order structure feature constraints. For the first two items, they are about the research on the improvement of the accuracy of RFM parameters and SAR imagery's error compensation model, by which the precise positioning of the multi-sensor images is obtained. For the third item, it realizes the robust corresponding points matching by introducing the RFM bundle adjustment and hypergraph matching techniques. This study is propitious to complementary advant
英文关键词: multi-sensor remote sensing imagery;spatial information extraction;rational function model;bundle adjustment;image matching