A new 2D depth-depth matching algorithm whose translation and rotation freedoms are separated

In this paper, we revise a previous 2D depth-depth-matching algorithm in order to copy motions fast from a real liver to a virtual liver in a surgical navigation. The real liver is always captured by 3D depth camera, and the virtual liver is represented by a polyhedron with STL format via DICOM captured by MRI/CT. In our algorithm, we firstly compare a 2D depth image in a real world and the Z-buffer of STL in a virtual world, and by using the difference of two depth images, we secondly search the best movement of a virtual liver from a huge number of possibilities with 3 translation and 3 rotation degrees-of-freedom. In this paper, we firstly divide translation and rotation D.O.F, and individually select the most adequate 3 DOF sets of a virtual liver following its real liver. Based on the division, we can find a sequence of following motions more precise and faster than our previous 2D depth-depth-matching algorithms.