3D Visual Servo Control of a Flexible Endoscope with RCM Constraint

In minimally invasive surgery, endoscopes provide visual guidance to surgical operations. To tackle the problems caused by traditional hand-held endoscope, robots are introduced to steer the endoscope. Image based instrument tracking is an effective way to ease the humanrobot-interaction. In this article, we aim to design a 3D visual servoing scheme for a robotic flexible surgical endoscope with algorithm-based RCM constraint. The endoscope consists of two identical cameras and is based on continuum mechanism, mounted on the tip of a UR5 robot arm. This article conducts kinematic modeling of the flexible robotic endoscope to get the mapping from joint space to task space. Then the relationship between depth and joint variables is derived. Finally, to validate the constrained visual servoing scheme, simulation is conducted in V-REP. The results show that it is feasible and effective to achieve 3D visual servoing of the flexible robotic endoscope with RCM constraints.