Marker-Based Instrument Tracking System Using Binocular Vision

The optical tracking system has rapidly developed into an essential part of infrared surgical navigation due to its high accuracy and convenience. This paper designs a high-precision and low-cost optical tracking system for preoperative training, which consists of a binocular-vision system, surgical tools, and image processing equipment. Based on the binocular vision principle, markers on the surgical instrument can be acquired by this system. To obtain the sub-pixel coordinates of the center of these markers' projections in real-time, first, a local image distortion correction algorithm was used to correct the region of interest in the captured image, next, the correct contours of the markers' projections can be acquired by filtering the extracted contours with geometric information, and then the accurate coordinates of the center of these markers' projections in pixel coordinate system can be obtained by leastsquares ellipse fitting algorithm. Finally, three different tests were performed on surgical instruments to verify the performance of the system, in which the stability test demonstrated that the system had high stability, as well as the static positioning test and dynamic tracking test proved that the system had high accuracy.