The use of Kinect sensor to control manipulator with electrohydraulic servodrives

Abstract

The article describes the control of the 2-axis electro- hydraulic manipulator by the human-hand motion. To recognition of skeleton points the Kinect sensor was used. In this application the information about coordinates of shoulder, elbow and hand was used to compute of inverse kinematic in manipulator. In investigation the accuracy of control by human's hand motion was tested. The aim of study was to find a new of control method without commonly used joysticks to create human-machine inter- face. owadays, many devices like robots, lifts, cranes and manipulators are designed to allow the easier control for human operator. There are performed wide range of research to develop ways of the human - machine commu- nication. The typical control system is based on the differ- ent kind of joysticks e.g. haptic joystick. In these structure there were also applies vision systems to analyze human's motion. Obtained data can be used to control mobile ro- bots or manipulators. These movement are also natural for human. Described systems can be potentially used to help in everyday life old people or for sick and disabled persons by e.g. provide medicaments closed them. They used Kinect sensor to recognition human motion and control the movement of the electric humanoid robot. The Human-Computer interaction was described by Vladimi (6). They proposed the computer vision-based analysis for interpretation of hand gestures to interaction human with computer. Fukuda (7) proposed a new master-slave ma- nipulator control system without mechanical master con- troller. A person whose forearm has been amputated can use this manipulator as a personal assistant for the desk- top work. A human-assisting manipulator was teleoperated by electromyographic (EMG) signals and arm motions. The use of kinect system for better quality of controlled process for the patient's rehabilitation was described by Fernandez-Baena (8). The investigation was based on com- parison markerless system with the Kinect sensor and expensive vision system. This work proved that the dispar- ity is not big between Kinect unit and expensive vision system. Therefore this system can be used to supervision motion of human. Nowadays communication between human-machine is very important. Many investigation are directed to im- prove control of machines by human's gestures. In this case programming of trajectory in machine may be easier rather than with use of some programming languages.