Indirect camera calibration for surgery tracking

Embedding cameras in surgical lights allows full view of surgery for tracking and storage purposes. However, it presents the challenge of camera calibration, as camera lights frequently change position. In this paper, we propose a novel indirect calibration method that utilizes static ceiling cameras that detect the position of light cameras and calibrate their extrinsic parameters every time there is significant movement. In our experimental setup, indirect calibration was able to locate light cameras with an average error of 4.81 mm of the position found by standard camera calibration methods and offers the added advantage of avoiding interference with the surgeons.