基于单相机的空间目标相对位姿测量系统

支帅; 丁国鹏; 韩世豪; 张永合; 朱振才

doi:10.37188/CO.2025-0057

基于单相机的空间目标相对位姿测量系统

doi: 10.37188/CO.2025-0057

cstr: 32171.14.CO.2025-0057

支帅^{1, 2, 3, ,},
丁国鹏^{1, 2,},
韩世豪^{1, 2},
张永合^{1, 2},
朱振才¹

1.
中国科学院微小卫星创新研究院, 上海 201304
2.
中国科学院卫星数字化技术重点实验室, 上海 201210
3.
中国科学院大学, 北京 100049

基金项目: “十四五”国家重点研发计划（No. 2024YFB3909400）；

详细信息

作者简介:
支　帅（1989—），女，辽宁锦州，博士研究生，2012年于湖南大学获得本科学位，2015年于中国科学院长春光学精密机械与物理研究所获得硕士学位，现于中国科学院微小卫星创新研究院，主要从事智能测量系统研究等相关工作。E-mail：zhis@microsate.com

中图分类号: TP394.1;TH691.9
计量
- 文章访问数: 0
- HTML全文浏览量: 0
- PDF下载量: 0
- 被引次数: 0
出版历程
- 网络出版日期: 2025-07-09

Monocular camera-based relative pose measurement system for space targets

ZHI Shuai^{1, 2, 3
, ,},
DING Guo-peng^{1, 2
,},
HAN Shi-hao^{1, 2},
ZHANG Yong-he^{1, 2},
ZHU Zhen-cai¹

1.
Innovation Academy for Microsatellites of Chinese Academy of Sciences, Shanghai 201304
2.
Key Laboratory of Satellite Digitization Technology, CAS, Shanghai 201210
3.
University of Chinese Academy of Sciences, Beijing 100049

Funds: Supported by

More Information

Corresponding author: zhis@microsate.com

摘要

摘要:
为提高测量系统的稳定性及精度，实现航天器超近距离高精度对接，本文提出了一种基于单相机及合作靶标的相对位姿测量系统，用于双星间相对位置及姿态的高精度测量。通过设计追踪星视觉相机及目标星LED合作靶标，在双星距离为50米到0.4米的范围内，实现了高精度的相对位姿测量。首先，通过设计的远近场LED靶标，实现了相机与靶标间的协同工作，保证在50米到0.4米的距离均能清晰成像；其次，根据设计的靶标特性，提出了多尺度质心提取算法，利用斜率一致性约束与间距比筛选，在复杂光照下稳定获取特征目标；最后，结合靶标几何约束的初值估计，实现了目标星相对于追踪星的位姿解算，为进一步提高测量精度，引入非线性优化方法对位姿结果进行迭代优化，有效降低了测量误差。试验结果表明，系统测量精度由远及近逐渐提高，在距离为0.4米时，位置测量精度优于1毫米，姿态测量精度优于0.2度，满足超近距离对接任务需求。本方案为空间在轨目标相对位姿测量提供了高精度、高稳定性的技术支撑，具有重要的工程应用价值。
- 单相机 /
- LED合作靶标 /
- 多尺度质心提取 /
- 非线性优化 /
- 相对位姿测量
Abstract:
To enhance measurement stability and accuracy for ultra-close-range target docking, this study proposes a monocular camera-based relative position and attitude measurement system with cooperative targets, enabling high-precision position and attitude determination between CubeSats. Through co-designed vision cameras on the chaser satellite and LED targets on the target satellite, precise relative pose measurement is achieved within 0.4−50 meters. First, the collaborative camera-target operation across far/near fields ensures clear imaging throughout the range from 50 meters to 0.4 meters. Second, a multi-scale centroid extraction algorithm incorporating slope consistency constraints and spacing ratio screening reliably acquires features under complex lighting conditions. Finally, combined with the initial pose estimation of the target satellite relative to the chaser, nonlinear optimization iteratively refines pose results to minimize errors. Experimental results demonstrate progressive accuracy improvement with proximity. At 0.4 m distance, the position measurement accuracy is better than 1 millimeter, and the attitude measurement accuracy is better than 0.2 degrees, satisfying ultra-close docking requirements. This solution provides high-precision, high-stability technical support for on-orbit space target relative navigation with important engineering application value.
- monocular camera /
- LED cooperative targets /
- multi-scale centroid extraction /
- non-linear optimization /
- relative pose measurement

HTML全文

图 1 测量相机及坐标系定义

Figure 1. Measurement camera and coordinate frame definition

下载: 全尺寸图片幻灯片

图 2 LED靶标及坐标系定义

Figure 2. LED targets and coordinate frame definition

下载: 全尺寸图片幻灯片

图 3 针孔成像模型

Figure 3. Pinhole Optics model

下载: 全尺寸图片幻灯片

图 4 图像坐标系

Figure 4. Image coordinate frame

下载: 全尺寸图片幻灯片

图 5 交叉投影法原理图

Figure 5. Schematic diagram of cross-projection method

下载: 全尺寸图片幻灯片

图 6 质心提取算法流程图

Figure 6. Flowchart of centroid extraction algorithm

下载: 全尺寸图片幻灯片

图 7 航天器间相对位姿示意图

Figure 7. Conceptual diagram of inter-spacecraft relative pose

下载: 全尺寸图片幻灯片

图 8 相对距离解算示意图

Figure 8. Diagram of relative range solution

下载: 全尺寸图片幻灯片

图 9 相对位姿解算算法流程图

Figure 9. Flowchart of relative pose solution algorithm

下载: 全尺寸图片幻灯片

图 10 光学成像组件实物图

Figure 10. Photograph of optical imaging assembly

下载: 全尺寸图片幻灯片

图 11 靶标组件实物图

Figure 11. Photograph of cooperative target assembly

下载: 全尺寸图片幻灯片

图 12 光学成像组件与靶标组件装星图

Figure 12. Flight integration diagram of optical imaging assembly and cooperative target assembly

下载: 全尺寸图片幻灯片

图 13 多尺度质心提取算法结果

Figure 13. Output of multi-scale centroid extraction algorithm

下载: 全尺寸图片幻灯片

图 14 50 m相对位置与姿态数据分布图

Figure 14. Relative position and attitude distribution map at 50 m range

下载: 全尺寸图片幻灯片

图 15 40 m相对位置与姿态数据分布图

Figure 15. Relative position and attitude distribution map at 40 m range

下载: 全尺寸图片幻灯片

图 16 20 m相对位置与姿态数据分布图

Figure 16. Relative position and attitude distribution map at 20 m range

下载: 全尺寸图片幻灯片

图 17 10 m相对位置与姿态数据分布图

Figure 17. Relative position and attitude distribution map at 10 m range

下载: 全尺寸图片幻灯片

图 18 5.6 m相对位置与姿态数据分布图

Figure 18. Relative position and attitude distribution map at 5.6 m range

下载: 全尺寸图片幻灯片

图 19 1.7 m相对位置与姿态数据分布图

Figure 19. Relative position and attitude distribution map at 1.7 m range

下载: 全尺寸图片幻灯片

图 20 0.4 m相对位置与姿态数据分布图

Figure 20. Relative position and attitude distribution map at 0.4 m range

下载: 全尺寸图片幻灯片

表 1 50 m相对位置与姿态测量结果

Table 1. Relative position and attitude measurement results at 50 m range

Parameters	Std	Mean-value	Max-v	Min-v
T-x(mm)	1.2015	−1013.5044	−1010.9075	−1016.7402
T-y(mm)	3.9683	−276.5156	−267.5542	−285.0395
T-z(mm)	52.8736	−50780.0293	−50652.7305	−50832.2031
$\varphi $ (°)	0.081	−0.05188	−0.0176	−0.0672
$\theta $ (°)	0.0154	0.07840	0.1064	0.0238
$\psi $(°)	0.0010	−0.0046	−0.0027	−0.0079

下载: 导出CSV

表 2 40 m相对位置与姿态测量结果

Table 2. Relative position and attitude measurement results at 40 m range

Parameters	Std	Mean-value	Max-v	Min-v
T-x(mm)	0.4432	−808.2581	−806.8610	−809.6720
T-y(mm)	0.2770	297.0678	297.9350	296.2810
T-z(mm)	19.9998	−40649.4954	−40590.0000	−40699.5000
$\varphi $ (°)	0.0050	−0.0845	−0.0651	−0.0972
$\theta $ (°)	0.0100	0.0680	0.1074	0.0351
$\psi $(°)	0.0015	−0.0055	−0.0019	−0.0097

下载: 导出CSV

表 3 20 m相对位置与姿态测量结果

Table 3. Relative position and attitude measurement results at 20 m range

Parameters	Std	Mean-value	Max-v	Min-v
T-x(mm)	0.3180	−283.5990	−283.2110	−285.5420
T-y(mm)	0.2050	245.5650	246.7930	244.7530
T-z(mm)	4.3640	−20730.6000	−20715.2000	−20742.5000
$\varphi $ (°)	0.0107	−0.0645	−0.0052	−0.0827
$\theta $ (°)	0.0108	0.0595	0.0807	−0.0091
$\psi $(°)	0.0011	−0.0079	−0.0054	−0.0111

下载: 导出CSV

表 4 10 m相对位置与姿态测量结果

Table 4. Relative position and attitude measurement results at 10 m range

Parameters	Std	Mean-value	Max-v	Min-v
T-x(mm)	0.7960	−193.6940	−190.0220	−194.9690
T-y(mm)	0.6060	209.2280	210.2260	208.2550
T-z(mm)	4.9600	−10329.8520	−10324.0660	−10345.1570
$\varphi $ (°)	0.0051	−0.0665	−0.0508	−0.0798
$\theta $ (°)	0.0055	0.0584	0.0670	−0.0420
$\psi $(°)	0.0006	−0.0065	−0.0041	−0.0079

下载: 导出CSV

表 5 5.6 m相对位置与姿态测量结果

Table 5. Relative position and attitude measurement results at 5.6 m range

Parameters	Std	Mean-value	Max-v	Min-v
T-x(mm)	0.1450	36.2800	36.5910	36.0940
T-y(mm)	0.9140	118.6200	119.2630	117.8860
T-z(mm)	1.6290	−5583.1500	−5576.1830	−5591.8940
$\varphi $ (°)	0.0230	−0.0080	0.0716	−0.0676
$\theta $ (°)	0.0250	−0.0020	0.1047	−0.0843
$\psi $(°)	0.0011	−0.0149	−0.0128	−0.0192

下载: 导出CSV

表 6 1.7 m相对位置与姿态测量结果

Table 6. Relative position and attitude measurement results at 1.7 m range

Parameters	Std	Mean-value	Max-v	Min-v
T-x(mm)	0.2270	11.8730	12.4530	11.6760
T-y(mm)	0.2810	65.3370	65.7950	64.9990
T-z(mm)	0.1870	−1654.8800	−1654.3430	−1655.4240
$\varphi $ (°)	0.0020	−0.0032	0.0078	−0.0029
$\theta $ (°)	0.0036	−0.0691	−0.0792	−0.0574
$\psi $(°)	0.0002	−0.0197	−0.0192	−0.0201

下载: 导出CSV

表 7 0.4 m相对位置与姿态测量结果

Table 7. Relative position and attitude measurement results at 0.4 m range

Parameters	Std	Mean-value	Max-v	Min-v
T-x(mm)	0.08	−58.63	−58.3670	−58.72
T-y(mm)	0.003	45.41	45.478	45.40
T-z(mm)	0.10	−405.79	−402.922	−405.89
$\varphi $ (°)	0.003	−0.041	0.0009	−0.043
$\theta $ (°)	0.006	−0.184	−0.046	−0.07
$\psi $(°)	0.0003	−0.0157	−0.007	−0.0161

下载: 导出CSV

参考文献(16)

[1]	ZHANG H F, WU J X, LIU D L, et al. Research on rocket engine pose measurement technology based on monocular vision[J]. Proceedings of SPIE, 2023, 12934: 129340I.
[2]	SEO C T, KANG S W, CHO M. Three-dimensional free view reconstruction in axially distributed image sensing[J]. Chinese Optics Letters, 2017, 15(8): 081102. doi: 10.3788/COL201715.081102
[3]	HEATON A F, HOWARD R T, PINSON R M. Orbital express AVGS validation and calibration for automated rendezvous[C]. AIAA/AAS Astrodynamics Specialist Conference and Exhibit, AIAA, 2008. (查阅网上资料, 未找到本条文献页码, 请确认).
[4]	KAWANO I, MOKUNO M, KASAI T, et al. Result of autonomous rendezvous docking experiment of engineering test Satellite-VII[J]. Journal of Spacecraft and Rockets, 2001, 38(1): 105-111. doi: 10.2514/2.3661
[5]	YAN K, XIONG ZH, LAO D B, et al. Attitude measurement method based on 2DPSD and monocular vision[J]. Proceedings of SPIE, 2019, 11338: 113382L.
[6]	MAO J F, HUANG W, SHENG W G. Target distance measurement method using monocular vision[J]. IET Image Processing, 2020, 14(13): 3181-3187. doi: 10.1049/iet-ipr.2019.1293
[7]	屈也频, 刘坚强, 侯旺. 单目视觉高精度测量中的合作目标图形设计[J]. 光学学报,2020,40(13):1315001. doi: 10.3788/AOS202040.1315001 QU Y P, LIU J Q, HOU W. Graphics design of cooperative targets on monocular vision high precision measurement[J]. Acta Optica Sinica, 2020, 40(13): 1315001. (in Chinese). doi: 10.3788/AOS202040.1315001
[8]	董永英, 张高鹏, 常三三, 等. 一种基于单目视觉的空间目标位姿测量算法及其精度定量分析[J]. 光子学报,2021,50(11):1112003. doi: 10.3788/gzxb20215011.1112003 DONG Y Y, ZHANG G P, CHANG S S, et al. A pose measurement algorithm of space target based on monocular vision and accuracy analysis[J]. Acta Photonica Sinica, 2021, 50(11): 1112003. (in Chinese). doi: 10.3788/gzxb20215011.1112003
[9]	RONDAO D, HE L, AOUF N. AI-based monocular pose estimation for autonomous space refuelling[J]. Acta Astronautica, 2024, 220: 126-140. doi: 10.1016/j.actaastro.2024.04.003
[10]	SANSONE F, FRANCESCONI A, OLIVIERI L, et al. Low-cost relative navigation sensors for miniature spacecraft and drones[C]. 2015 IEEE Metrology for Aerospace (MetroAeroSpace), IEEE, 2015: 389-394.
[11]	PIRAT C, ANKERSEN F, WALKER R, et al. Vision based navigation for autonomous cooperative docking of CubeSats[J]. Acta Astronautica, 2018, 146: 418-434. doi: 10.1016/j.actaastro.2018.01.059
[12]	BUI M T, DOSKOCIL R, KRIVANEK V. Distance and angle measurement using monocular vision[C]. 2018 18th International Conference on Mechatronics, IEEE, 2018: 1-6.
[13]	陈天择, 葛宝臻, 罗其俊. 重投影优化的自由双目相机位姿估计方法[J]. 中国光学,2021,14(6):1400-1409. doi: 10.37188/CO.2021-0105 CHEN T Z, GE B ZH, LUO Q J. Pose estimation for free binocular cameras based on reprojection error optimization[J]. Chinese Optics, 2021, 14(6): 1400-1409. (in Chinese). doi: 10.37188/CO.2021-0105
[14]	CAPUANO V, KIM K, HARVARD A, et al. Monocular-based pose determination of uncooperative space objects[J]. Acta Astronautica, 2020, 166: 493-506. doi: 10.1016/j.actaastro.2019.09.027
[15]	ZHANG ZH, BIN W, KANG J H, et al. Dynamic pose estimation of uncooperative space targets based on monocular vision[J]. Applied Optics, 2020, 59(26): 7876-7882. doi: 10.1364/AO.395081
[16]	PIAZZA M, MAESTRINI M, DI LIZIA P. Monocular relative pose estimation pipeline for uncooperative resident space objects[J]. Journal of Aerospace Information Systems, 2022, 19(9): 613-632. doi: 10.2514/1.I011064