A computed stereoscopic method for laparoscopic surgery by using feature tracking

This paper presents a technique that generates rough depth maps from laparoscopy camera using feature tracking techniques for depth sampling. The method is applicable to single laparoscopy video images through local motion cues to improve the inferred depth maps. The feature tracking is used to ensure temporal depth consistency, such as SURF. The results show that the proposed methods significantly improve the depth estimation for laparoscopy surgery. The technique can be used to convert a 2D laparoscopy video into stereo for 3D visualization, and it demonstrates this through a laparoscopy surgery video, including results of surgical instruments.