A Gesture Model and Its Applications in Risk Assessment for Misrecognition

Before the gesture recognition is to be widely used as a new method for human-robot interface, some measurements are required for risk assessment and for risk reduction to assure safety, especially if the interface is to be used in action control for the robot working in human-robot cooperation style. In order to assess and to reduce the risk from misrecognition of gestures, a quantitative 3D model was proposed for representing gestures, and a quantitative measurement, called gesture distance, was proposed for the evaluation about how much the difference between gestures are. As the example of using the gesture distance, the method to implement the gesture recognition based on the gesture distance was briefly described, and then the risk assessment for a gesture recognition system was given based on the anthropometric data in Japanese Anthropometric Database.