Abstract
Sensor systems have been developed that can automatically obtain physiological data in bathrooms and toilet stalls. In order to measure the physiological data of individuals who live with a number of other people, each person has to be identified separately. However, it is difficult to obtain active user cooperation. Ideally, personal identification should be achieved without requiring subjects to cooperate. Therefore, we measured changes in the load and trajectory of the center of foot pressure when opening a door by utilizing the features of a Nintendo® Wii balance board as a load sensor. We calculated the Euclidean distance between the registered data and the test data. Personal identification was achieved by using the nearest neighbor algorithm. Data generated while walking through a door were obtained for five subjects, yielding an identification rate of 93.7%.
