Gesture recognition based on wearable sensing

With the continuous development of virtual reality technology, data gloves have become an important means of virtual exchange. Aiming at the defect, this paper proposes a kind of wearable controller based on inertial sensor with gesture recognition function. The controller hardware includes inertial sensors composed by accelerometer and gyro, WIFI communication module and Microprocessor. The gesture recognition algorithm based on the extraction of acceleration and angular velocity feature, the decision tree classifier was carried out on the two kinds of similar gesture recognition, and considering the change of attitude angle to further distinguish gestures category. The experimental results show that gesture recognition average accuracy of the wearable controller is 95.98%, which can well realize the gesture recognition function.