A sequence-action recognition applying state machine for user interface

The user interface in accelerometer-enable mobile devices requires more advanced action recognition technology than that of the traditional input method due to its awkward handling (e.g., limited computational capability, miniaturized input/output controls, and so on). Many types of input technology and interaction styles are being tested for gesture recognition. We introduce a method based on state machines that constructs high-level interaction events from lower-level ones. This is a novel user interface that recognizes a user's sequence-action by using the two-axis accelerometer in order to supplement the lacking user interface of mobile devices. To recognize a user's actions (user's gesture and posture) in the order of precedence, we used a state machine algorithm. We designed mobile games such as fishing and bowling to be played on mobile phones by using this algorithm. We will prove that this state machine is appropriate for sequence-action recognition.