Visual servoing in a cable robot using Microsoft Kinect v2 sensor

Tracking the 3D pose of an object plays an important role in many applications. However, it has been a challenging problem in visual servoing systems due to high-cost sensors and requiring some invasive changes for the object or environment. This paper is focused on tracking an object using the low-cost Microsoft Kinect sensor. The goal is tracking an object with a parallel cable robot's moving platform (MP) simulator by the Kinect sensor. Two different methods by using Kinect sensor will be introduced. The first is based on a colorbase technique (Camshift) in MATLAB, while the second exploits the color and depth information using OpenCV library. Finally, these methods are experimentally implemented to generate the desired path for the parallel cable robot simulator.