Camera-IMU extrinsic calibration based on prior poses and motion planning

doi:10.11947/j.AGCS.2026.20250287

Abstract

Abstract:

Multi-sensor fusion leverages the complementarity of heterogeneous data to achieve high-precision navigation and positioning, where accurate sensor extrinsic calibration serves as a fundamental prerequisite. To address the limitations of traditional camera-IMU calibration methods—such as strong reliance on calibration targets, sensitivity of calibration performance to data quality, and cumbersome data acquisition—this paper proposes a targetless calibration algorithm based on prior poses and motion planning. The method utilizes GNSS/SINS post-processed smoothed trajectories as prior poses to construct a reprojection error model, and employs a Gauss-Newton optimization framework to estimate the extrinsic parameters with high accuracy. A motion planning strategy is developed to optimize the data acquisition trajectory, using a motion turntable to sufficiently excite the IMU and enhance image overlap, thereby ensuring the repeatability and reliability of the calibration data. Furthermore, a complete initialization and extrinsic refinement pipeline is introduced to accelerate convergence and achieve optimal calibration results. Simulation results demonstrate that the proposed method achieves an extrinsic calibration accuracy better than 0.05° in orientation and 1 cm in translation. Field experiments further validate the feasibility and effectiveness of the proposed calibration approach and workflow.

Key words: integrated navigation, sensor calibration, inertial navigation system, multi-sensor fusion

CLC Number:

P228

Rui ZHOU, Feng ZHU, Xiaohong ZHANG. Camera-IMU extrinsic calibration based on prior poses and motion planning[J]. Acta Geodaetica et Cartographica Sinica, 2026, 55(3): 465-476.

Figures/Tables 17

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Tab. 1

Tab. 2

Fig. 6

Tab. 3

Fig. 7

Tab. 4

Fig. 8

Fig. 9

Fig. 10

Tab. 5

Fig. 11

Tab. 6

References 23

[1]	杨元喜. 导航与定位若干注记[J]. 导航定位学报, 2015, 3(3): 1-4.
	YANG Yuanxi. Notes of navigation and positioning[J]. Journal of Navigation and Positioning, 2015, 3(3): 1-4.
[2]	辜声峰, 戴春齐, 何成鹏, 等. 面向城市车载导航的多系统PPP-RTK/VIO半紧组合算法性能分析[J]. 武汉大学学报(信息科学版), 2021, 46(12): 1852-1861.
	GU Shengfeng, DAI Chunqi, HE Chengpeng, et al. Analysis of semi-tightly coupled multi-GNSS PPP-RTK/VIO for vehicle navigation in urban areas[J]. Geomatics and Information Science of Wuhan University, 2021, 46(12): 1852-1861.
[3]	朱锋. GNSS/SINS/视觉多传感器融合的精密定位定姿方法与关键技术[D]. 武汉: 武汉大学, 2019.
	ZHU Feng. GNSS/SINS/vision multi-sensors integration for precise positioning and orientation determination[D]. Wuhan: Wuhan University, 2019.
[4]	龚宁静. 无人系统多传感器标定问题研究[D]. 哈尔滨: 哈尔滨工业大学, 2021.
	GONG Ningjing. Research on multi-sensor calibration for unmanned system[D]. Harbin: Harbin Institute of Technology, 2021.
[5]	黄平, 胡超, 张宁, 等. 车载LiDAR-IMU外参联合标定算法[J]. 仪器仪表学报, 2022, 43(10): 128-135.
	HUANG Ping, HU Chao, ZHANG Ning, et al. Vehicle-mounted LiDAR-IMU external parameter joint calibration algorithm[J]. Chinese Journal of Scientific Instrument, 2022, 43(10): 128-135.
[6]	张守建, 李欣然, 王逸石, 等. 城市环境下视觉惯性里程计辅助PPP定位[J]. 导航定位学报, 2023, 11(6): 34-41, 101.
	ZHANG Shoujian, LI Xinran, WANG Yishi, et al. Visual inertial odometry-aided PPP positioning in urban environments[J]. Journal of Navigation and Positioning, 2023, 11(6): 34-41, 101.
[7]	姚二亮, 张合新, 张国良, 等. 基于Vision-IMU的机器人同时定位与地图创建算法[J]. 仪器仪表学报, 2018, 39(4): 230-238.
	YAO Erliang, ZHANG Hexin, ZHANG Guoliang, et al. Robot simultaneous localization and mapping algorithm based on vision and IMU[J]. Chinese Journal of Scientific Instrument, 2018, 39(4): 230-238.
[8]	QIN Tong, LI Peiliang, SHEN Shaojie. VINS-mono: a robust and versatile monocular visual-inertial state estimator[J]. IEEE Transactions on Robotics, 2018, 34(4): 1004-1020.
[9]	CAMPOS C, ELVIRA R, RODRÍGUEZ J J G, et al. ORB-SLAM3: an accurate open-source library for visual, visual-inertial, and multimap SLAM[J]. IEEE Transactions on Robotics, 2021, 37(6): 1874-1890.
[10]	LEUTENEGGER S, LYNEN S, BOSSE M, et al. Keyframe-based visual-inertial odometry using nonlinear optimization[J]. The International Journal of Robotics Research, 2015, 34(3): 314-334.
[11]	LÜ Jiajun, ZUO Xingxing, HU Kewei, et al. Observability-aware intrinsic and extrinsic calibration of LiDAR-IMU systems[J]. IEEE Transactions on Robotics, 2022, 38(6): 3734-3753.
[12]	潘林豪, 田福庆, 应文健, 等. 单目相机-IMU外参自动标定与在线估计的视觉-惯导SLAM[J]. 仪器仪表学报, 2019, 40(6): 56-67.
	PAN Linhao, TIAN Fuqing, YING Wenjian, et al. VI-SLAM algorithm with camera-IMU extrinsic automatic calibration and online estimation[J]. Chinese Journal of Scientific Instrument, 2019, 40(6): 56-67.
[13]	LI Mingyang, MOURIKIS A I. High-precision, consistent EKF-based visual-inertial odometry[J]. The International Journal of Robotics Research, 2013, 32(6): 690-711.
[14]	LI Mingyang, YU Hongsheng, ZHENG Xing, et al. High-fidelity sensor modeling and self-calibration in vision-aided inertial navigation[C]//Proceedings of 2014 IEEE International Conference on Robotics and Automation. New Yrok: IEEE, 2014: 409-416.
[15]	SCHNEIDER T, LI Mingyang, CADENA C, et al. Observability-aware self-calibration of visual and inertial sensors for ego-motion estimation[J]. IEEE Sensors Journal, 2019, 19(10): 3846-3860.
[16]	QIN Tong, SHEN Shaojie. Online temporal calibration for monocular visual-inertial systems[C]//Proceedings of 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems. New Yrok: IEEE, 2019: 3662-3669.
[17]	YANG Zhenfei, SHEN Shaojie. Monocular visual-inertial state estimation with online initialization and camera-IMU extrinsic calibration[J]. IEEE Transactions on Automation Science and Engineering, 2017, 14(1): 39-51.
[18]	YANG Yulin, GENEVA P, ECKENHOFF K, et al. Degenerate motion analysis for aided INS with online spatial and temporal sensor calibration[J]. IEEE Robotics and Automation Letters, 2019, 4(2): 2070-2077.
[19]	MIRZAEI F M, ROUMELIOTIS S I. A Kalman filter-based algorithm for IMU-camera calibration: observability analysis and performance evaluation[J]. IEEE Transactions on Robotics, 2008, 24(5): 1143-1156.
[20]	FURGALE P, BARFOOT T D, SIBLEY G. Continuous-time batch estimation using temporal basis functions[C]//Proceedings of 2012 IEEE International Conference on Robotics and Automation. New Yrok: IEEE, 2012: 2088-2095.
[21]	OTH L, FURGALE P, KNEIP L, et al. Rolling shutter camera calibration[C]//Proceedings of 2013 IEEE Conference on Computer Vision and Pattern Recognition. New Yrok: IEEE, 2013: 1360-1367.
[22]	MAYE J, FURGALE P, SIEGWART R. Self-supervised calibration for robotic systems[C]//Proceedings of 2013 IEEE Intelligent Vehicles Symposium (IV). New Yrok: IEEE, 2013: 473-480.
[23]	CHEN Shuolong, LI Xingxing, LI Shengyu, et al. Targetless spatiotemporal calibration for multiple heterogeneous cameras and IMUs based on continuous-time trajectory estimation[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 9515112.

先验位姿误差	外参标定误差	RMS	2σ值
0.02 m/0.1°（2像素）	旋转x轴/（°）	0.015	0.032
	旋转y轴/（°）	0.019	0.039
	旋转z轴/（°）	0.020	0.041
	平移x轴/m	0.002	0.004
	平移y轴/m	0.002	0.004
	平移z轴/m	0.003	0.006

先验位姿误差		像素误差/像素	RMS		解算成功率/（%）
位置/m	姿态/（°）	像素误差/像素	旋转外参/（°）	平移外参/m	解算成功率/（%）
0.00	0.1	2	0.004	0.001	100
0.01			0.009	0.001	100
0.02			0.018	0.002	100
0.03			0.025	0.004	99.2
0.04			0.038	0.006	96.1
0.05			0.050	0.009	92.5
0.02	0.0	2	0.017	0.002	100
	0.1		0.018	0.002
	0.2		0.018	0.002
	0.3		0.021	0.003
	0.4		0.026	0.003
	0.5		0.030	0.004
0.02	0.1	0	0.017	0.002	100
		1	0.017	0.002
		2	0.018	0.002
		3	0.018	0.002
		4	0.019	0.002
		5	0.019	0.002

先验位姿误差	像素误差/像素	远近点阈值/m	RMS		解算成功率/（%）
位置/m姿态/（°）	像素误差/像素	远近点阈值/m	旋转外参/（°）	平移外参/m	解算成功率/（%）
0.020.1	2	1	0.027	0.006	100
		2	0.025	0.006
		3	0.022	0.004
		4	0.018	0.002
		5	0.018	0.002
		6	0.019	0.003
		7	0.020	0.005
		8	0.023	0.005
		9	0.024	0.006
		10	0.024	0.009

传感器类型	典型参数	参数值
MEMS-IMU（HGuide I300）	速度随机游走
	角度随机游走
	加表零偏不稳定性	20 μg
	陀螺零偏不稳定性	3（°）/h
	观测频率	200 Hz
单目相机（MER-131）	快门类型	全局快门
	分辨率	1280×1024像素
	观测频率	10 Hz
GNSS接收机（Mosaic-X5）	GNSS天线	HX-CH7609A
GNSS接收机（Mosaic-X5）	观测频率	10 Hz

标定方法	旋转外参/（°）	平移外参/m
设计值	（－90.00，0.00，180.00）	（0.150，0.000，－0.140）
本文方法	（－89.61，－0.19，179.63）	（0.144，－0.007，－0.134）
Kalibr	（－90.96，－0.32，179.11）	（0.139，0.005，－0.128）