中国科学院沈阳自动化研究所机构知识库
Advanced  
SIA OpenIR  > 光电信息技术研究室  > 学位论文
题名: 空间非合作目标视觉位姿测量关键技术研究
其他题名: Research on Key Technologies for Vision Pose Measurement of Space Non-cooperative Target
作者: 苗锡奎
导师: 朱枫
分类号: TP391.41
关键词: 计算机视觉 ; 非合作目标 ; 位姿定位 ; 典型部件 ; 天地协同 ; 多相机非共视场 ; 鲁棒性
页码: 113页
学位专业: 模式识别与智能系统
学位类别: 博士
答辩日期: 2013-05-20
授予单位: 中国科学院沈阳自动化研究所
作者部门: 光电信息技术研究室
中文摘要: 自主接近和近距离及超近距离作业是重要的空间在轨任务,要执行这样的空间任务,需要测量目标之间的相对位姿。基于视觉的方法是位姿测量的重要方法之一,国内外在空间合作目标视觉位姿测量方面已经取得了大量的研究成果,而且已经有很多具体工程应用。相对来说,非合作目标视觉位姿测量研究的理论和成果还比较薄弱,还没有相关的工程应用。本文以交会对接和机械手近距离自主作业为背景,开展面向空间非合作目标的视觉位姿测量关键技术研究,具有重要的理论意义和现实意义。论文分别从位姿估计方法、典型部件测量方法、天地协同测量方法和多相机非共视场测量方法四个方面对视觉位姿测量开展研究,定位于应用基础研究和技术攻关,为工程应用提供理论支持和技术支撑,研究内容主要包括以下四个方面: 1. 根据特征点噪声为各向异性且非独立同分布,而且常常具有很强的方向不确定性的特点,提出一种基于特征点不确定性加权误差的位姿估计方法,该方法利用协方差矩阵对特征点方向不确定性建模,并将方向不确定性融入到重投影误差中,构造基于不确定性加权误差的目标函数,最后利用Levenberg-Marquardt算法优化目标函数求解位姿。实验结果表明,该方法可以适应不同程度的方向不确定性,精度优于现有迭代方法,而且随着不确定性的增加,位姿解的精度并没有明显变差。2. 根据空间飞行器上具有视觉特征和几何特征已知的典型部件的特点,选择利用太阳能帆板部件与星箭对接环部件完成视觉位姿测量。前者根据矩形帆板的视觉特征设计了专用的特征提取算法,可自动提取帆板上的门形三线和外侧角点,然后利用点特征或线特征算法解算目标位姿;后者根据对接环所提供的单圆特征,针对单圆定位存在二义性问题,将对接环平面外一参考点到圆心距离的欧式不变性作为约束,提出一种唯一确定对接环姿态和位置参数的方法。实验结果表明,两种方案都能有效地确定飞行器的位置与姿态。3. 针对视觉特征和几何特征不完全已知的空间飞行器,无法自动完成特征提取与位姿解算问题,提出天地协同的位姿测量思想,将星载相机所拍摄的目标图像下传至地面,由地面操作人员上传合作信息辅助完成特征提取与位姿解算。基于该思想完成了基于飞行器主体四边的天地协同测量方案和基于姿控发动机基座的天地协同测量方案。两种方案都使用天地协同的方式辅助完成矩形四边的提取,利用矩形四角点特征解算位姿。为了提取测量目标的直线段特征,提出一种方向预测搜索与边缘曲率约束的直线段提取方法,以为位姿解算提供更准确有效的数据。实验结果表明,天地协同的位姿测量能有效解决特征不完全已知的空间飞行器位姿测量问题。4. 为解决近距离下,成像空间小、特征尺寸比较大,单相机或多目立体相机很难对目标上的特征完整成像,而无法完成定位问题。在目标为点特征的情况下设计了一种多相机非共视场位姿迭代求解方法,该方法利用不同视角的多个相机可以得到目标更完整的信息,同时也可以利用目标上的冗余信息,以换取测量系统对噪声的鲁棒性,进一步提高了位姿解算算法的综合性能。在此基础上给出了多相机非共视场位姿测量系统中位姿误差的表示,从误差传递的角度,采用简化条件下定性几何分析和直接仿真实验验证的方法分析测量位姿对特征点图像坐标检测误差的鲁棒性与三种特征点几何构形的关系,得到了光轴方向上和光轴垂直方向上分布的特征点以及测量距离对各位姿量的影响规律。得出的结论可以指导目标上特征点构形的选择,从而达到优化系统设计,提高整个定位系统性能的目的,对于视觉位姿测量系统的工程应用具有一定的意义。
英文摘要: Autonomous approaching and operation in close and ultra-close range are important space on-orbit missions, to carry out such space missions, measuring the relative pose between two targets is indispensable. Vision-based approach is one of the important methods of pose measurement. Researchers have made lots of achievements in vision pose measurement of space cooperative target at home and abroad. The theories and achievements of vision pose measurement of space non-cooperative target are still weak. The dissertation takes rendezvous and docking and manipulator autonomous operation as research background, the research on key technologies for vision pose measurement of space non-cooperative target is carried out, and has important theoretical and practical significance. The study on vision pose measurement is carried out mainly from four aspects: pose estimation method, typical component oriented measurement method, collaboration of space-ground measurement method and multiple non-overlapping cameras measurement method, aiming at applied basic research and technology research. The main contents and results obtained are as follows: 1. According to the fact that measurement noises are not only anisotropic and non-independent and non-identically distributed, but also have strong directional uncertainty, a new pose estimation method is proposed. Firstly, the directional uncertainty of feature points is modeled using covariance matrix, and incorporated into reprojection errors, and then the uncertainty-weighted cost function is constructed. Finally, we use Levenberg-Marquardt method to optimize the cost function to get the pose. Experimental results indicate that the method can adapt to varying degrees of directional uncertainty and guarantee more accurate results, and the method does not degrade with increase in directionality of uncertainty. 2. According to typical components on spacecrafts, solar panel component and docking ring component oriented vision pose measurement schemas are proposed. In the former, rectangular feature can be extracted automatically by specially designed algorithm according to the visual feature of the solar panel, employing high-precision and robust algorithm to calculate the pose. In the latter, pose measurement based on single circle suffers from two solutions. In order to exclude the interference of false solution, a new false solution removal method using euclidean distance invariance as a constraint is proposed to solve the pose ambiguity. Experimental results indicate that both schemas can effectively determine the pose of the spacecrafts. 3. Aiming at spacecrafts whose visual feature and geometric constraints are not completely known, pose measurement based on collaboration of space-ground is proposed. The basic idea is that downloading the image captured by the camera in spacecraft through telemetry channel, operators in the ground upload artificially cooperative information to the satellite program, and the satellite program will easily complete feature extraction and pose calculation. Based on the idea, collaboration of space-ground pose measurement schemas based on rectangle of main body of spacecraft and attitude control engine rectangular base are proposed. In order to extract lines feature of the target, a new line segments detection method based on direction prediction searching and curvature constraints is proposed. Experimental results indicate that collaboration of space-ground pose measurement can effectively solve the pose problem of spacecrafts whose features are not completely known. 4. To solve the problem that monocular and multi-view stereo cameras can not complete the pose measurement in close and ultra-close range, a new pose measurement method using multiple cameras with non-overlapping views is proposed. The method can not only get more complete information about the target, but also takes advantage of the redundant information to make the measurement system more robustness to the noise. From the perspective of error propagation and though qualitative geometric analysis under simple conditions and direct simulation experiment under common conditions, we make a overall analysis on the relationship between the robust performance of pose parameters which relate to image coordinate detection error and three types of feature points geometric configuration. The influence law of pose parameters affected by the feature points on the optical axis direction, the vertical direction of the optical axis and the measuring distance is drawn. The conclusion can guide the choice of feature points geometric configuration on the target, so as to optimize the design of measurement system and improve the overall performance of the measurement system, and will also be of great significance in the engineering application of pose measurement system.
语种: 中文
产权排序: 1
内容类型: 学位论文
URI标识: http://ir.sia.cn/handle/173321/10765
Appears in Collections:光电信息技术研究室_学位论文

Files in This Item:
File Name/ File Size Content Type Version Access License
空间非合作目标视觉位姿测量关键技术研究.pdf(1963KB)----限制开放 联系获取全文

Recommended Citation:
苗锡奎.空间非合作目标视觉位姿测量关键技术研究.[博士学位论文].中国科学院沈阳自动化研究所.2013
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[苗锡奎]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[苗锡奎]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2016  中国科学院沈阳自动化研究所 - Feedback
Powered by CSpace