An Acceleration Feature-Based Gesture Recognition System

With the diversification of the sensing element that was built-in smart mobile device, it has become one of the important devices in people's daily life. Among these sensors, the accelerometer is the most common sensor that was embedded in smart mobile device. Therefore, this study proposed a gesture recognition algorithm based on the acceleration feature. It utilized the fuzzy control technique to classify different gestures. The proposed gesture recognition system provides a low complexity and high accuracy gesture recognition method. The simulation results show that the proposed method can recognize these gestures that include tilted to the left side, turn to the other side, shaking, and Z-shaped and have the accuracy rate of more than 91%. In future works, the system can be applied in health care systems or smart home applications to provide intuitive manipulation via specific gesture.