A system stores clustering information obtained by clustering multiple sets of viewpoints relative to teeth of a three-dimensional model of a patient and projection images of a head including the teeth from the respective viewpoints. At the time of a surgery, the system creates comparative images from a photographed image taken by a monocular camera and searches for any viewpoint having a projection image similar to the comparative image sequentially from a first hierarchy to a final hierarchy in the clustering information. The system sets a representative viewpoint based on the viewpoint of the final hierarchy, restores position and posture of the teeth of the patient relative to the monocular camera as position and posture of the teeth of the three-dimensional model relative to the representative viewpoint, and generates and outputs a navigation image based on position and posture of surgical information relative to the teeth after the restoration.
