Selection of Appliance Using Skeletal Tracking and 3D Face Tracking for Gesture Control Home Automation

Abstract

This study implements 3d face tracking and skeletal tracking in Microsoft Kinect Xbox One. Once the user is detected by the sensor head point from skeletal tracking, and a computed point D from chin point and eyebrow midpoint from 3d face tracking will be used to create the Line of Sight vector. Also, appliance points are always specified by the location of the appliance in the room with respect to the Kinect. Different appliance vectors will be created through vector subtraction. Angles between the Line of Sight vector and each of the appliance vector were computed through scalar product and compared to obtain the smallest angle. Once the smallest angle was obtained it was compared to a 15-degree threshold. If it’s within the threshold, then the appliance is selected.