Loading...

Integration of the inertial navigation system with consecutive images of a camera by relative position and attitude updating

Ghanbarpour Asl, H ; Sharif University of Technology | 2019

683 Viewed
  1. Type of Document: Article
  2. DOI: 10.1177/0954410019852818
  3. Publisher: SAGE Publications Ltd , 2019
  4. Abstract:
  5. This paper introduces a new method for improving the inertial navigation system errors using information provided by the camera. An unscented Kalman filter is used for integrating the inertial measurement unit data with the features’ constraints extracted from the camera’s image. The constraints, in our approach, comprise epipolar geometry of two consecutive images with more than 65% coverage. Tracking down a known feature in two consecutive images results in emergence of stochastic epipolar constraint. It emerges in the form of an implicit measurement equation of the Kalman filter. Correctly matching features of the two images is necessary for reducing the navigation system errors because they act as external information for the inertial navigation system. A new method has been presented in this study based on the covariance analysis of the matched feature rays’ intersection points on the ground, which sieves the false matched features. Then, the inertial navigation system and matched feature information is integrated through the unscented Kalman filter filter and the states of the vehicle (attitude, position, and velocity) are corrected according to the last image. In this paper, the relative navigation parameters against the absolute one have been corrected. To avoid increasing dimensions of the covariance matrix, sequential updating procedure is used in the measurement equation. The simulation results show good performance of the proposed algorithm, which can be easily utilized for real flights. © IMechE 2019
  6. Keywords:
  7. Epipolar constraint ; Feature tracking ; Inertial navigation system ; Integrated navigation system ; Visual navigation ; Air navigation ; Cameras ; Covariance matrix ; Flight simulators ; Image processing ; Kalman filters ; Matched filters ; Navigation ; Stochastic systems ; Epipolar constraints ; Feature-tracking ; Inertial measurement unit ; Integrated navigation systems ; Navigation system errors ; Relative position and attitude ; Unscented Kalman Filter ; Inertial navigation systems
  8. Source: Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering ; Volume 233, Issue 15 , 2019 , Pages 5592-5605 ; 09544100 (ISSN)
  9. URL: https://journals.sagepub.com/doi/10.1177/0954410019852818