Gesture Control of a Robotic Head using Kinect

Abstract

Robots, from the simplest to the most complex, require remote control. Teleoperation requires contact, whether a joystick, a controller, or an HMI is used. In most cases, these devices do not offer the necessary comfort, nor the possibility for the teleoperator to be able to use his skills and abilities at an optimal level. Many studies have been done for sensors to solve this discomfort, some of which also offer wearable sensors such as: microwave multiple-input multiple-output (MIMO) arrays, IMU sensors, fisheye monocular camera for human posture detection. In this paper, he proposes a 3D posture estimation solution that uses a Kinect V1 device to fuse RGB and depth images. We developed a MatLab Simulink implementation algorithm to automatically compute angles of both shoulder joints, elbow joints, and wrist joints based on a skeleton model that allows us to estimate 20 joint positions. To validate our experiment, we offer a non-contact solution that can control the joints of a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$3\mathrm{D}$</tex> printed robotic head based only on user right-hand movements.