基于序列图像的深空探测精确定点着陆自主导航方法

Autonomous navigation based on sequence image is the key technology of pinpoint landing mission for future deep space and is the main way of deep space exploration technology. For the key difficulty problems in the autonomous navigation based on sequence image for pinpoint landing, this project studies the autonomous navigation theory based on sequence images for pinpoint landing on deep space celestial body, with emphasis on science problems of multi-dimensional and multi-scale image information representation, feature extraction and matching，and accurate fusion navigation. This project will try to overcome key problems in autonomous navigation for pinpoint landing on deep space celestial body. The results have widely potential applications in pinpoint landing missions on large celestial bodies, such as the Moon, and the Mars, and on small celestial bodies, such as asteroids and comets. The applicant has won one second-class of State Technological Innovation Award, three first-class of National Defense Invention Award and Science and Technology Progress Awards. In addition, the applicant has licensed patents for more than 40 copies, authored one monographs, and published more than 40 papers.

基于序列图像的自主导航是未来深空探测精确定点着陆任务必需的关键技术，是深空探测技术的重点发展方向之一。针对基于序列图像的深空探测精确定点着陆自主导航存在的难点问题，项目重点研究天体表面地形信息的稀疏表征、多尺度多视点图像特征提取与匹配以及多信息最优融合导航处理等科学问题，提出导航路标的稀疏表征与优化选取、图像特征的鲁棒提取与匹配以及精确着陆的多信息最优融合导航等方法，从理论方法层面突破深空天体精确着陆自主导航的核心技术，项目研究成果可以直接应用与月球、火星等大天体以及小行星、彗星等小天体着陆探测任务，具有广泛的应用前景。项目负责人在相关领域已取得的主要成果有：获国家技术发明二等奖1项，国防发明和进步一等奖3项，获授权发明专利40多项，专著1部，发表论文40多篇。

基于序列图像的自主导航是未来深空探测精确定点着陆任务必需的关键技术，是深空探测技术的重点发展方向之一。针对基于序列图像的深空探测精确定点着陆自主导航存在的难点问题，项目重点研究了天体表面地形信息的稀疏表征、多尺度多视点图像特征提取与匹配以及多信息最优融合导航处理等关键问题，提出了导航路标的稀疏表征与优化选取、图像特征的鲁棒提取与匹配以及精确着陆的多信息最优融合导航等方法，形成了基于序列图像的精确定点着陆自主导航方法，构建了序列图像自主导航地面仿真验证系统。项目已发表论文17篇，出版专著2本（中文版和英文版、部分内容），申请发明专利4项。项目研究成果可以直接应用于月球、火星等大天体以及小行星、彗星等小天体着陆探测任务，具有广泛的应用前景，已应用在我国无人月球后续、载人登月、火星采样返回以及小天体探测等任务论证中。