空间目标序列图像细粒度检测仿真数据集设计

doi:10.16182/j.issn1004731x.joss.25-1073

系统仿真学报 ›› 2026, Vol. 38 ›› Issue (9): 2711-2724.doi: 10.16182/j.issn1004731x.joss.25-1073

• 论文 • 上一篇

空间目标序列图像细粒度检测仿真数据集设计

王世妍¹^,²^,³^,⁴, 赵旦谱¹^,²^,³^,⁴, 洪海昆¹^,²^,³^,⁴

^1.中国科学院空天信息创新研究院，北京 100190
^2.中国科学院空间信息处理与应用系统技术重点实验室，北京 100190
^3.中国科学院目标认知与应用技术重点实验室，北京 100190
^4.中国科学院大学电子电气与通信工程学院，北京 100049

收稿日期:2025-11-04 修回日期:2025-12-05 出版日期:2026-09-30 发布日期:2026-10-02
通讯作者: 赵旦谱
第一作者简介:王世妍(2000-)，女，博士生，研究方向为空间目标仿真与识别、深度学习。

Design of a Simulation Dataset for Fine-grained Detection of Space Target Sequence Images

Wang Shiyan¹^,²^,³^,⁴, Zhao Danpu¹^,²^,³^,⁴, Hong Haikun¹^,²^,³^,⁴

^1.Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, China
^2.Key Laboratory of Technology in Geo-spatial Information Processing and Application System, Chinese Academy of Sciences, Beijing 100190, China
^3.Key Laboratory of Target Cognition and Application Technology (TCAT), Chinese Academy of Sciences, Beijing 100190, China
^4.School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

Received:2025-11-04 Revised:2025-12-05 Online:2026-09-30 Published:2026-10-02
Contact: Zhao Danpu

摘要/Abstract

摘要：

随着卫星等空间目标数量的不断增加，远距离空间目标的细粒度检测与识别需求日益迫切。然而，实测数据获取成本高、标注粒度不足等问题制约了相关研究的发展。针对上述问题，构建基于多模块协同的空间目标序列图像仿真系统，通过STK和Blender联合仿真，模拟空间目标图像生成场景，自动输出序列仿真图像和对应的细粒度标签。基于上述系统，构建空间目标序列图像细粒度检测仿真数据集，包括12类空间目标图像数据和3类部件的标签数据。并利用该数据集进行空间目标分类和细粒度检测实验，生成基准结果，供相关学者参考使用。该数据集为地面端的模型训练和算法验证提供了标准化数据，推动空间目标检测技术的发展，并为后续自主监测和智能决策任务奠定基础。

关键词: 空间目标, 仿真数据集, 序列图像, 细粒度检测, 恒星背景

Abstract:

With the increasing number of space objects such as satellites, there is a growing need for fine-grained detection and recognition of distant space targets. However, obtaining real observational data remains difficult due to high costs and insufficient annotation detail. To overcome these limitations, asynthetic dataset generation system based on multi-module collaboration is developed. Using joint simulation with STK and Blender, the proposed system simulates optical imaging conditions of space targets and automatically produces sequential images along with corresponding fine-grained labels. A large-scale simulation dataset is constructed for fine-grained detection in space target image sequences. It contains images of 12 categories of space targets and includes component-level labels for 3 types of key parts. Experiments are carried out on space target classification and fine-grained component detection, providing baseline results for relevant research to use. This dataset supports model training and algorithm verification in ground-based applications. It promotes the development of space target detection technology and lays a foundation for autonomous space monitoring and intelligent decision-making.

Key words: space target, simulation dataset, sequence image, fine-grained detection, star background

中图分类号:

TP391.4

王世妍,赵旦谱,洪海昆 . 空间目标序列图像细粒度检测仿真数据集设计[J]. 系统仿真学报, 2026, 38(9): 2711-2724.

Wang Shiyan,Zhao Danpu,Hong Haikun . Design of a Simulation Dataset for Fine-grained Detection of Space Target Sequence Images[J]. Journal of System Simulation, 2026, 38(9): 2711-2724.

图/表 18

图1

图2

表1

STK仿真参数设置

参数名		参数值
仿真场景参数	仿真总时长/h	24
	仿真步长/s	6
	传感器追踪模式	Targeted
观测相机参数	相机视场角 $θ F O V$	60°×60°
	焦距/mm	450
	可观测最大视星等	15

表1

图3

图4

图5

图6

图7

图8

图9

图10

图11

表2

表3

图12

图13

图14

表4

目标检测实验结果

算法	类别	Precision	Recall	$m A P @ 50$	$m A P @ 50 - 95$
YOLOv5	all	0.771	0.672	0.748	0.447
	satellite	0.946	0.787	0.866	0.559
	panel	0.639	0.580	0.634	0.351
	body	0.729	0.650	0.745	0.431
YOLOv8	all	0.935	0.632	0.740	0.542
	satellite	0.934	0.886	0.944	0.763
	panel	0.938	0.382	0.523	0.334
	body	0.934	0.629	0.754	0.529
YOLOv11	all	0.897	0.670	0.760	0.523
	satellite	0.924	0.839	0.898	0.714
	panel	0.868	0.438	0.572	0.333
	body	0.900	0.734	0.812	0.521
YOLOv13	all	0.858	0.631	0.726	0.523
	satellite	0.866	0.897	0.950	0.774
	panel	0.848	0.362	0.495	0.309
	body	0.858	0.634	0.732	0.486

表4

参考文献 28

[1]	陈思伟, 李铭典, 崔兴超. Pol-ISARSpaceTarget-1.0: 极化ISAR空间目标精细识别电磁仿真数据集[J]. 信号处理, 2025, 41(8): 1443-1454.
	Chen Siwei, Li Mingdian, Cui Xingchao. Pol-ISARSpaceTarget-1.0: Polarimetric ISAR Electromagnetic Simulation Dataset for Space Target Fine Recognition[J]. Journal of Signal Processing, 2025, 41(8): 1443-1454.
[2]	Massimi Federica, Ferrara Pasquale, Petrucci Roberto, et al. Deep Learning-based Space Debris Detection for Space Situational Awareness: A Feasibility Study Applied to the Radar Processing[J]. IET Radar, Sonar & Navigation, 2024, 18(4): 635-648.
[3]	Stokes G H, Curt von Braun, Sridharan R, et al. The Space-based Visible Program[C]//Space 2000 Conference and Exposition. Reston: AIAA, 2000: AIAA 2000-5334.
[4]	吴郯. 基于深度学习的空间目标智能检测与识别算法研究[D]. 西安: 西安电子科技大学, 2020.
	Wu Tan. Research on Intelligent Detection and Recognition Algorithm of Space Target Based on Deep Learning[D]. Xi'an: Xidian University, 2020.
[5]	Yan Zhenguo, Song Xin, Zhong Hanyang. Spacecraft Detection Based on Deep Convolutional Neural Network[C]//2018 IEEE 3rd International Conference on Signal and Image Processing (ICSIP). Piscataway: IEEE, 2018: 148-153.
[6]	Wu Tan, Yang Xi, Song Bin, et al. T-SCNN: A Two-stage Convolutional Neural Network for Space Target Recognition[C]//IGARSS 2019—2019 IEEE International Geoscience and Remote Sensing Symposium. Piscataway: IEEE, 2019: 1334-1337.
[7]	Afshar Roya, Lu Shuai. Classification and Recognition of Space Debris and Its Pose Estimation Based on Deep Learning of CNNs[C]//HCI International 2020 - Posters. Cham: Springer International Publishing, 2020: 605-613.
[8]	Priyadarshini I. Enhanced Space Debris Detection and Monitoring Using a Hybrid Bi-LSTM-CNN and Bayesian Optimization[J]. Artificial Intelligence and Applications, 2025, 3(1): 43-55.
[9]	Yuan Man, Zhang Guhong, Yu Zhuoqun, et al. Spacecraft Components Detection Based on a Lightweight YOLOv3 Model[C]//2022 IEEE 10th Joint International Information Technology and Artificial Intelligence Conference (ITAIC). Piscataway: IEEE, 2022: 1968-1973.
[10]	Tang Qiang, Li Xiangwei, Xie Meilin, et al. Intelligent Space Object Detection Driven by Data from Space Objects[J]. Applied Sciences, 2024, 14(1): 333.
[11]	Chen Yulang, Gao Jingmin, Zhang Kebei. R-CNN-based Satellite Components Detection in Optical Images[J]. International Journal of Aerospace Engineering, 2020(1): 8816187.
[12]	Wang Chien-Yao, Bochkovskiy Alexey, Liao Hongyuan. YOLOv7: Trainable Bag-of-freebies Sets New State-of-the-art for Real-time Object Detectors[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2023: 7464-7475.
[13]	张浩鹏, 刘正一, 姜志国, 等. BUAA-SID1.0空间目标图像数据库[J]. 航天返回与遥感, 2010, 31(4): 65-71.
	Zhang Haopeng, Liu Zhengyi, Jiang Zhiguo, et al. BUAA-SID1.0 Space Object Image Dataset[J]. Spacecraft Recovery & Remote Sensing, 2010, 31(4): 65-71.
[14]	Proença Pedro F, Gao Yang. Deep Learning for Spacecraft Pose Estimation from Photorealistic Rendering[C]//2020 IEEE International Conference on Robotics and Automation (ICRA). Piscataway: IEEE, 2020: 6007-6013.
[15]	Kisantal Mate, Sharma S, Park T H, et al. Satellite Pose Estimation Challenge: Dataset, Competition Design, and Results[J]. IEEE Transactions on Aerospace and Electronic Systems, 2020, 56(5): 4083-4098.
[16]	Park T H, Märtens Marcus, Lecuyer Gurvan, et al. SPEED+: Next-generation Dataset for Spacecraft Pose Estimation across Domain Gap[C]//2022 IEEE Aerospace Conference (AERO). Piscataway: IEEE, 2022: 1-15.
[17]	Mohamed Adel Musallam, Kassem Al Ismaeil, Oyedotun Oyebade, et al. SPARK: SPAcecraft Recognition Leveraging Knowledge of Space Environment[EB/OL]. (2021-04-14) [2025-10-26]. .
[18]	Hoang Anh Dung, Chen Bo, Chin Tat-Jun. A Spacecraft Dataset for Detection, Segmentation and Parts Recognition[C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Piscataway: IEEE, 2021: 2012-2019.
[19]	Perez M D, Mohamed Adel Musallam, Garcia Albert, et al. Detection & Identification of On-Orbit Objects Using Machine Learning[C]//8th European Conference on Space Debris. Darmstadt: ESA Space Debris Office, 2021: 1-10.
[20]	李鹏飞, 徐伟, 朴永杰, 等. 天基空间小目标复杂场景数字成像仿真[J]. 系统仿真学报, 2025, 37(9): 2315-2334.
	Li Pengfei, Xu Wei, Yongjie Piao, et al. Digital Imaging Simulation of Complex Scene of Space-based Space Small Target[J]. Journal of System Simulation, 2025, 37(9): 2315-2334.
[21]	张健, 娄树理, 任建存. 一种空间观测图像的仿真生成方法[J]. 电光与控制, 2014, 21(11): 18-23.
	Zhang Jian, Lou Shuli, Ren Jiancun. A Simulation Method for Space Observation Image[J]. Electronics Optics & Control, 2014, 21(11): 18-23.
[22]	袁麟, 吕品, 郑昌文, 等. 星空环境成像效果的仿真研究[J]. 系统仿真学报, 2009, 21(8): 2174-2178, 2184.
	Yuan Lin, Pin Lü, Zheng Changwen, et al. Research on Simulation of Imaging Effects in Star Field[J]. Journal of System Simulation, 2009, 21(8): 2174-2178, 2184.
[23]	姚保利, 雷铭, 薛彬, 等. 高分辨和超分辨光学成像技术在空间和生物中的应用[J]. 光子学报, 2011, 40(11): 1607-1618.
	Yao Baoli, Lei Ming, Xue Bin, et al. Progress and Applications of High-resolution and Super-resolution Optical Imaging in Space and Biology[J]. Acta Photonica Sinica, 2011, 40(11): 1607-1618.
[24]	Cortes C, Vapnik V. Support-vector Networks[J]. Machine Learning, 1995, 20(3): 273-297.
[25]	Breiman L. Random Forests[J]. Machine Learning, 2001, 45(1): 5-32.
[26]	Cover T, Hart P. Nearest Neighbor Pattern Classification[J]. IEEE Transactions on Information Theory, 1967, 13(1): 21-27.
[27]	He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep Residual Learning for Image Recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2016: 770-778.
[28]	Lei Mengqi, Li Siqi, Wu Yihong, et al. YOLOv13: Real-time Object Detection with Hypergraph-enhanced Adaptive Visual Perception[EB/OL]. (2025-09-05) [2025-10-26]. .

类目	详细信息
CPU	8核Ryzen 7 6800H 3.2 GHz
GPU	NVIDIA GeForce RTX 3050 Laptop GPU
内存	16 GByte
硬盘	512 GByte
操作系统	Windows 11 家庭中文版
系统类型	x64-based PC
编程语言	Python
编程工具	Visual Studio Code

算法	Accuracy	Precision	Recall	F1
SVM	0.676	0.695	0.676	0.683
随机森林	0.490	0.470	0.490	0.475
MLP	0.479	0.471	0.492	0.471
逻辑回归	0.638	0.634	0.638	0.633
KNN	0.256	0.312	0.256	0.266
ResNet	0.444	0.520	0.444	0.418

空间目标序列图像细粒度检测仿真数据集设计

Design of a Simulation Dataset for Fine-grained Detection of Space Target Sequence Images

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 18

参考文献 28

相关文章 5

编辑推荐

Metrics

本文评价

[1]	张扬, 甘庆波, 袁洪, 傅江良, 孙成明. 空间目标天基监视分布式仿真系统设计[J]. 系统仿真学报, 2020, 32(4): 620-626.
[2]	林浩申, 杨晓君, 何兵. 空间目标模糊平方根容积卡尔曼滤波定位算法[J]. 系统仿真学报, 2018, 30(4): 1361-1368.
[3]	杨振发, 万刚, 曹雪峰, 李锋, 谢理想. 基于几何结构特征的点云表面重建方法[J]. 系统仿真学报, 2017, 29(11): 2684-2692.
[4]	施群山, 蓝朝桢, 梁静, 卢万杰, 王栋. 空间目标多尺度表达模型及方法[J]. 系统仿真学报, 2016, 28(9): 2235-2245.
[5]	马跃龙, 赵勇, 曹雪峰. 一种基于单目SLAM的无人机序列图像拼接方法[J]. 系统仿真学报, 2016, 28(10): 2349-2355.