项目名称: 星表遥感图像优化及典型地物精细刻画方法研究
项目编号: No.U1331108
项目类型: 联合基金项目
立项/批准年度: 2014
项目学科: 数理科学和化学
项目作者: 尹继豪
作者单位: 北京航空航天大学
项目金额: 60万元
中文摘要: 随着我国月球及深空探测技术的不断发展,星表遥感数据的应用成为关注的热点。受地形和成像条件的限制,获取的星表遥感图像,特别是极区图像,不能满足实际应用的需求,因此遥感图像优化显得尤为重要。课题针对极区影像优化、高空间分辨率图像和高光谱图像融合两方面开展图像优化工作,并应用优化后的图像对撞击坑等典型地物进行精细刻画。 具体研究内容包括:研究极区图像阴影去除与信息恢复方法;研究利用时间序列分析方法提取图像特征的方法,并在此基础上设计相应的融合策略;研究压缩感知理论在高空间分辨率图像和高光谱图像融合中的具体应用;基于生成的星表优化数据,开展星表典型地物,如撞击坑的形态识别、地形分析、成分探测等工作,完成对撞击坑的精细刻画。关键技术的理论突破与实现将有助于充分利用嫦娥系列卫星传回的数据,实现我国探月工程制定的科学目标,并为火星等深空探测计划做好前期准备。
中文关键词: 压缩感知理论;阴影检测与去除;图像融合;光谱解混;撞击坑描述
英文摘要: In recent years, China’s Chang-E series probing satellite has completed the task of moon exploration successfully. The application of remote sensing data has become the focus of space science research. Restricted by terrain and illumination conditions, the acquired planetary remote sensing data, such as the image data of polar region and hyperspectral remote sensing data, cannot satisfy the need of actual application. Therefore, it is very important for the optimization of remote sensing image. Focusing on image optimization of polar region and hyperspectral image fusion, we plans to study multi-source image fusion method of planet surface. Based on the optimization data, we will purse the accurate description of craters. Firstly, we consider the algorithms of shadow elimination and information recovery on polar image. Secondly, using the time series analysis theory, we discuss the feature extraction algorithm in hyperspectral and panchromatic images. At the same time, we study the specific applications of compressed sensing theory in hyperspectral image fusion. Based on the research above, we intend to design the corresponding fusion strategy. Thirdly, using the optimized data, we conduct shape recognition, terrain analysis, composition detection in impact craters to accomplish their accurate description. Brea
英文关键词: Compressed sensing;Shadow detection and removal;Image fusion;Unmixing;Crater characterization