Can a Display Be Used for Rapid LiDAR-Camera Extrinsic Calibration?

Key points

• The camera observes displayed AprilTags, whereas the LiDAR measures the physical screen surface and boundary, requiring indirect cross-modal association.1
• Display geometry introduces association ambiguities, while indoor clutter complicates extraction.1
• ScreenCalib addresses these challenges through tag-guided display extraction, geometric hypothesis generation and refinement, and surrounding scene edges for ambiguity resolution.1
• The joint multi-pose mode combines independent static observations through cross-pose consensus and jointly refines the shared extrinsics and tag scale, reducing sensitivity to tag-size measurement errors without requiring continuous motion or inter-pose trajectory estimation.1

Sources (1)

• [1]Can a Display Be Used for Rapid LiDAR-Camera Extrinsic Calibration?
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 20, 05:29 PM

  ScreenCalib enables light detection and ranging (LiDAR)--camera extrinsic calibration using an ordinary programmable display, providing known visual markers and controlled geometry while reducing dedicated target preparation.

  The camera observes displayed AprilTags, whereas the LiDAR measures the physical screen surface and boundary, requiring indirect cross-modal association.

Extractive summary: sentences quoted from the sources.