A study for assistant robotic system using motion capture method for adapting to human-robot interface

In this paper, the authors present a user-centered concept for the operation experience of human-robot interface. For showing a concrete concept of our research, we design an assistant robot system. There are several stages for developing this system. Firstly, it is to obtain user´s data. The authors obtained static data from Internet, and activity data from capturing humans´ a motion data. The authors adapt accelerator sensor and image data, to capture human motion. Secondly, the authors use association rule of data mining technique to find the hiding knowledge from user´s data. The system can base on the found knowledge, which includes the relation about life patterns, psychological conditions and etc., to provide appropriate assistance. The system mainly includes the following steps: capture user´s motion data by accelerator sensor and stereo camera in indoor environment, analysis captured motion data, found the hiding knowledge from user´s data, recognize the environment through user´s data, and based on these data and knowledge to provide service. In this paper, we design a human-centered assistant system for adapting to human-robot interface. The system not only uses static data, but also activity data using capture methods. We briefly describe the assistant robotic system´s overall concept, and discuss the motion captures methods and knowledge finding. Simple experiment results will also be shown.