摘要
针对小天体附着过程中形貌特征复杂且数量众多,导致在线特征选取时特征组合数量多、计算量大、计算效率低等问题,提出了一种在线形貌特征快速选取方法。利用多类型特征快速评价指标结合搜索区域,建立了单帧图像特征快速优选的方法。同时,考虑到小天体附着过程中在线特征选取采样间隔短、帧间图像重合度高、特征继承概率高的特点,提出了帧间图像特征继承优选方法。仿真实验表明单帧图像搜索区域和帧间图像特征继承的设计,降低了特征组合数量,提升了计算效率,提高了在线特征选取速度。
Given the complex and numerous morphological features during asteroid landing,which lead to challenges such as huge number of feature combinations,high computational load in online feature selection and low computational efficiency,in this paper an online multi-type feature rapid selection method was proposed.Using rapid selection criteria for multi-type features combined with search region,a method for selecting single-frame image features was established.Additionally,considering the short sampling interval,high inter-frame image overlap,and high feature inheritance probability during asteroid landing,a method for inter-frame image feature inheritance selection was proposed.Simulation experiments show that the design of single-frame image search regions and inter-frame image feature inheritance methods substantially reduces the number of feature combinations,thereby greatly enhancing the efficiency of computational and online feature selection.
作者
修文博
何颖
朱圣英
刘延杰
XIU Wenbo;HE Ying;ZHU Shengying;LIU Yanjie(Institute of Deep Space Exploration,Beijing Institute of Technology,Beijing 100081,China;Key Laboratory of Autonomous Navigation and Control for Deep Space Exploration,Ministry of Industry and Information Technology,Beijing 100081,China;International Cooperation Center of China Aerospace,Beijing 100048,China;College of Automation&Electronic Engineering,Qingdao University of Science and Technology,Qingdao 266042,China)
出处
《深空探测学报(中英文)》
CSCD
北大核心
2024年第3期286-294,共9页
Journal Of Deep Space Exploration
基金
基础科研项目(JCKY2021603B030)。
关键词
小天体附着
多类型特征
特征快速选取
形貌特征
视觉导航
asteroid landing
multi-type feature
rapid feature selection
morphology feature
visual navigation