Joint Estimation of Radar Odometry and Radar Mounting Parameters
Year: 2026
Type: conference-paper
Abstract: This paper presents a joint estimation framework for radar odometry and radar mounting parameter estimation. The kinematic model of radar detections is expressed in terms of detection measurements, host vehicle speeds, and radar mounting parameters. Stationary detections are identified using a residual-based statistical gating method. Radar odometry is then estimated via least squares, while mounting parameters are estimated through nonlinear least squares. A joint solution is achieved using an alternating least squares (ALS) scheme. Results on real-world radar data show that the proposed algorithm improves accuracy, reduces data restrictions, and requires lightweight computation suitable for real-time deployment.
Authors: Jinhyoung Oh, Prateek Reddy Nallabolu
Institution: Lucid Motors (United States)
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7208763598