Method of Developing Hand Gesture Sets for Vehicle Entertainment and Ambiance Controls

Along with the development of autonomous and semi-autonomous driving, how people spend their time in vehicles will also change and these changes may require new methods of interacting with services in the vehicle such as entertainment, ambiance, and air conditioning. Based on these anticipated needs, this study investigates free air hand gesture control as a method of developing a set of hand gestures for control. The objective of this research is to model how people express commands to machine interfaces with free air hand gestures in order to develop consensus gesture sets. To collect the gesture data, participants perform hand gestures following 18 types of experiment tasks, and the gestures are recorded by 3 cameras. K-means++ clustering of the landmark coordinate data for every 5 frames extracts common gesture sequences for each task. The clustered sequences are then used to develop consensus gesture sets are developed for each task.