Recieved:

20/07/2026

Accepted:

12/10/2026

Page: 

–

doi:

http://dx.doi.org/10.17515/resm2026-1936vk0720rs

Views:

24

Motion image stabilization of vehicle-based Fourier telemetry system based on multi-source data fusion

Li Wan 1,2,3, Liang Xu2, Wenqing Liu2, Qiangling Zhang 2

1Science Island Branch, Graduate School of USTC, Hefei, China
2Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei, China
3Institute of Environment Hefei Comprehensive National Science Center, Hefei, China

Abstract

Fourier telemetry is essential for vehicle-mounted gas tracing and environmental monitoring. However, during dynamic motion, motion blur, sensor drift, and environmental interference severely decrease imaging stability. To deal with these challenges, a vehicle-mounted Fourier remote sensing motion image stabilization method based on multi-source data fusion was proposed. A collaborative framework is constructed to enable global geographic spatial constraints and local dynamic pose perception. This framework is integrated with high-precision GIS data, GPS, and visual-inertial observations. A sliding window-based joint pose optimization algorithm is developed to align estimated poses with ideal imaging poses in the East-North-Up (ENU) coordinate system. This algorithm can effectively suppress cumulative errors and achieve stable motion imaging. The proposed method innovatively incorporates GIS as a real-time global constraint, optimizing pose estimation and coordinate alignment in a staged manner. Also, this method integrates ORB feature matching with GIS coordinate verification to enhance robustness. From the experimental results, this system achieves a motion drift RMSE of 0.054 m, an attitude jitter RMSE of 5.976°, and an imaging instability ratio of only 2.00% in dynamic scenarios. The relationship between the 5.976° azimuth angle RMSE value and its axial components (0.03°, 0.10°, 0.11°) lies in that the azimuth angle RMSE is the combined result of the errors of the three axial components. Its value is mainly controlled by the larger errors in the pitch and roll directions, while the error in the yaw direction is relatively smaller. A high-quality stable imaging support is provided for vehicle-mounted Fourier telemetry in gas leakage detection and smart city environmental monitoring. To clarify the overall stabilization logic, Fig. 1 presents the imaging stabilization schematic diagram that illustrates the system’s basic workflow and key functional modules.

Keywords

Multi-Source data fusion; Vehicle-Mounted platform; Fourier telemetry; Motion image stabilization; Visual-inertial fusion; Sliding window optimization

Cite this article as: 

Share This Article
LinkedIn
X
Facebook
journal cover
News & Upcoming Events