Development of a Gesture Analysis System for Human System Interaction
Chifuyu Matsumoto, Yuka Sone, Jinseok Woo, Yasuhiro Ohyama · 2024
Recently, Japan has been facing various social issues, and to solve these problems, it is necessary to create a society where diverse people can live comfortably. To build such a society, appropriate support tailored to individuals is required, necessitating the analysis of various types of information to understand the people's needs at any given moment. In particular, analyzing non-verbal elements such as posture, facial expressions, and gestures is crucial, as these naturally indicate a person's condition. Therefore, this study focuses on human gestures, one of the non-verbal elements, and proposes a gesture recognition method using Dynamic Time Warping (DTW) and K-Means. We demonstrated that symbolic gesture recognition is possible by recording arm trajectories from skeletal measurements with an RGB-D camera.