Dynamic characteristic analysis of users' motions for human smartphone interface

摘要

The aim of this study is to investigate characteristics of user's hand-motion to control electronic devices using a smartphone as a remote controller. Existing motion recognition techniques are not easy to apply to the smartphone directly, since it has limited resources compared to the desktop or notebook computer. But, smartphones have an advantage over any other computer, since smartphones are equipped with various sensors including orientation sensor and acceleration sensor. Nowadays, smartphones have various input ways to communicate with user through text, image or QR code and user's motion. Also, smartphones are able to communicate and exchange data with other electronic devices using Wifi or Buletooth technique. It gives lots of possibilities to offer convenience to consumers and one of examples is smartphone can be used as an input devices or a remote controller for home appliances. It will be particularly helpful to elders or disabled persons since the phone is the closest to person than any other electronic device on the current human life style. To develop better application software as a remote controller operated by user's motion, the motion should be investigated and analyzed trajectories of smartphone to get dynamic characteristics of user's actions. In this research, three different periodic motions such as tilt, zooming in-out and swing were studied by analyzing trajectories of a smartphone using outputs of orientation sensor and acceleration sensor. The result of our experiment shows that the average one period time of three motions is 1.88±0.07sec while one period time for three motion are 1.84sec for tilt, 1.98sec for zoom in-out and 1.81sec for swing. The result of additional experiment of additional two motions shows that two motions' one period time are within the range of three motions'. At the current stage, the result shows that stereotype of hand motion's one gesture is about 1.9sec, but it needs further research to reconfi- m. However, the stereotype of hand motion's one gesture means that any repetitive hand motion using smartphone could be utilized as one of input way, such as motion of moving up-down and moving right-left. Furthermore, if the motion shows a periodic aspect, user does not need to wait until sensor is stabilized or calibrated at the initial operating stage. The result of this research will be applied to the universal design or barrier free design of home appliances to give better accessibility and convenience to everybody including elders and handicapped persons.