Image-Based Visual Servoing Using an Optimized Trajectory Planning Technique

Abstract

Trajectory planning is a useful technique in robotics for guiding the robot through complicated tasks. In this paper, a new semi-offline trajectory planning method is developed to perform image-based visual servoing (IBVS) tasks for a 6 DOFs robotic manipulator system. This method extends the operation range of the system compared with the traditional IBVS controllers. In this method, the camera's velocity screw is parametrized using time-based profiles. The parameters of the velocity profile are then determined by minimizing the cost function consisting of the error between the initial and desired features while respecting the system constraints. A depth-estimation algorithm is proposed to provide the trajectory planning algorithm with a good estimation of the initial depth. The algorithm for planning the orientation of the robot is decoupled from the position planning of the robot. This method eliminates the limitation caused by camera's field of view. The algorithm is validated via the experiment on a 6 DOFs Denso robot in an eye-in-hand configuration. The experimental results demonstrate that the proposed method can overcome some major IBVS drawbacks such as surpassing the system limits and causing instability of the system in fulfilling the tasks which require a 180° rotation of the camera about its center.