Mobile platform motion control system based on human gestures

Abstract

Controllers based on human gestures recognition may be a suitable and user-friendly solution for the interaction between machines and humans. The academic research on this topic may benefit the industry and the society in general, deserving a proper support from governmental institutions. A challenging field of application of such systems is the control of mobile platforms. Traditional mobile platform controllers may be limitative regarding freedom of movements and are inadequate in some applications. Mobile platforms with four Mecanum wheels allow the improvement of pre-existing solutions, such as wheelchairs or forklifts, endowing them with higher manoeuvrability, but they require a complex motion control because of the wide variety of movements allowed. The control of all possible movements of such a platform cannot be achieved with a simple joystick or steering wheel. This paper describes a motion control system for a mobile platform with four Mecanum wheels. The system is based on hand tracking and gestures recognition using an Intel RealSense 3D sensor.