Vehicle Trajectory Data-Based Orientation Calibration for Two Oppositely Installed Roadside Millimeter-Wave Radars: With and without Prior Road Alignment Knowledge

The cooperative operation of roadside millimeter-wave (MMW) radar sensors can effectively extend the detection range of a single unit through data splicing. However, accurate calibration of their positions is required as the foundation for data splicing. In this research, a novel framework is proposed, using trajectory data to calibrate the orientation of opposing roadside MMW radar. This approach takes advantage of overlapping detection regions to act as a “binding agent” in the calibration process. Considering practical constraints where prior alignment knowledge may be unavailable, two methods are developed that respectively mine radar data association through trajectory analysis for data splicing. These methods were applied to the roadside MMW radar data collected from the G94 highway. The calibrated trajectory plots confirm the successful establishment of data splicing between the opposing roadside MMW radar units using both proposed calibration methods. This research provides valuable insights for similar multi-sensor cooperative calibration tasks.