Wodtko, Thomas (Ulm University), Deuscher, Marco (Ulm University), Buchholz, Michael (Universität Ulm)

Leveraging Motion Tracking for Sample Weighting in Motion-Based Calibration and Self-Assessment

Scheduled for presentation during the Regular Session "S28b-Multi-Sensor Fusion and Perception for Robust Autonomous Driving" (TH-EA-T28), Thursday, November 20, 2025, 13:50−14:10, Stradbroke

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Abstract

Extrinsic sensor calibration is crucial for reliable and robust multi-sensor fusion in autonomous systems. While various approaches exist for obtaining such information for different sensor types, motion-based extrinsic calibration has proven to be best suited for online evaluation since it can be used in the environment in which an autonomous system is deployed. Nevertheless, once an extrinsic calibration is determined, it is subject to change due to external influences like temperature, vibrations, or even vandalism. Consequently, it must be assessed online to ensure the safe and robust operation of the autonomous system. Existing methods are only capable of recalculating an extrinsic calibration online; however, this may be computationally expensive and may potentially use corrupt data. In this work, we first present a weighting approach based on motion tracking. Then, leveraging the tracking results, we propose an online self-assessment approach for extrinsic calibration, including a certainty measure, which allows us to determine if recalibration is an option. Based on real-world motion patterns provided by the EuRoC MAV dataset, we evaluate our approach and demonstrate its superiority compared to state-of-the-art approaches. Our approach's applicability comprises various autonomous systems and, thus, contributes to the safe and robust operation of intelligent transportation systems.