Pose estimation of near-infrared surgical instruments based on binocular vision

To enhance the accuracy of surgical instrument pose estimation within the surgical navigation system, a near-infrared surgical instrument pose estimation system based on binocular stereoscopic vision was devised. Initially, near-infrared markers were affixed to the surgical instruments, and their positions were captured using binocular cameras. Leveraging the transformation between pixel and world coordinates, along with stereoscopic vision principles, the spatial positions of the markers were determined. Subsequently, these marker positions were utilized to establish the instrument coordinate system, aligned with the camera coordinate system. Finally, employing the motion model of the instrument in space and utilizing the least squares algorithm with the minimum mean square error criterion, the instrument calibration process was optimized. The Extended Kalman Filter (EKF) was then deployed to refine the dynamic pose estimation of the instrument. The experimental outcomes illustrated that the stability and precision of the instrument pose estimation met the demands of surgical navigation, with a root mean square error (RMSE) of 0.23 mm for dynamic tip position estimation and a maximum error of 2 degrees in the estimation of the instrument's orientation. This approach enables automatic, accurate, and stable estimation of surgical instrument poses.