Laparoscopic Image Analysis for Robotic Arm Guidance

摘要

Some of the most important problems arising during a laparoscopic surgery relate to the limited perception of depth and field of view (fov). In this paper we explore the possibilities of dealing with these problems using appropriate re-modeling of the laparoscopic equipment and image processing algorithms, as to increase the perception ability of the surgeon within the operation space. We demonstrate that by using two camera units and appropriate estimation of the epipolar geometry between the camera views, we can acquire an accurate 3D map of the scene, as well as increased field of view. The benefits for the surgeon include the enhancement of the visible operating space and the increased perception of this space, with the ability of accurately estimating distances of sensitive abdomen regions and organs from the end of the laparoscope tip. Several simulation and real-world experiments are presented as to validate the potential of the proposed scheme.