PLV-IEKF: Consistent Visual-Inertial Odometry using Points, Lines, and Vanishing Points (2311.04477v1)

Abstract: In this paper, we propose an Invariant Extended Kalman Filter (IEKF) based Visual-Inertial Odometry (VIO) using multiple features in man-made environments. Conventional EKF-based VIO usually suffers from system inconsistency and angular drift that naturally occurs in feature-based methods. However, in man-made environments, notable structural regularities, such as lines and vanishing points, offer valuable cues for localization. To exploit these structural features effectively and maintain system consistency, we design a right invariant filter-based VIO scheme incorporating point, line, and vanishing point features. We demonstrate that the conventional additive error definition for point features can also preserve system consistency like the invariant error definition by proving a mathematically equivalent measurement model. And a similar conclusion is established for line features. Additionally, we conduct an invariant filter-based observability analysis proving that vanishing point measurement maintains unobservable directions naturally. Both simulation and real-world tests are conducted to validate our methods' pose accuracy and consistency. The experimental results validate the competitive performance of our method, highlighting its ability to deliver accurate and consistent pose estimation in man-made environments.

• The paper introduces a right invariant IEKF-based VIO method that leverages points, lines, and vanishing points to enhance consistency and minimize angular drift.
• It employs mathematically equivalent error definitions for point and line features to preserve system consistency in man-made environments.
• Simulations and real-world experiments validate the approach, demonstrating competitive performance in accurate and robust pose estimation.

The paper "PLV-IEKF: Consistent Visual-Inertial Odometry using Points, Lines, and Vanishing Points" presents an innovative approach to Visual-Inertial Odometry (VIO), which is crucial for accurate navigation and localization in robotic and autonomous systems. The authors tackle common issues associated with conventional Extended Kalman Filter (EKF)-based VIO methods, specifically system inconsistency and angular drift. These issues arise naturally in environments where feature-based methods are employed for localization.

In response to these challenges, the authors introduce an Invariant Extended Kalman Filter (IEKF) scheme that leverages structural regularities found in man-made environments, such as lines and vanishing points. These features provide additional cues that can significantly enhance the accuracy and consistency of pose estimation.

The central innovation of the paper lies in the application of a right invariant filter-based VIO that incorporates three types of features:

• Points: The paper demonstrates that the conventional additive error definition for point features can preserve system consistency, similar to the invariant error definition. This is achieved by proving a mathematically equivalent measurement model.
• Lines: A similar approach is taken for line features, ensuring that the system's consistency is maintained.
• Vanishing Points: The authors conduct an invariant filter-based observability analysis, showing that vanishing point measurements inherently maintain unobservable directions. This characteristic is beneficial for maintaining the system's accuracy over time.

The paper includes both simulation and real-world experiments to validate the proposed method. The results indicate that the PLV-IEKF scheme delivers competitive performance, offering accurate and consistent pose estimation in structured environments. This achievement underscores the potential of utilizing multiple structural features to enhance the robustness of visual-inertial odometry systems in practical applications.