Safyzan Salim, Muhammad Mahadi Abdul Jamil, Radzi Ambar and Suraya Mohammad
Keywords: artificial neural network, hand gesture recognition, sign language translator sensory glove, EMG
Abstract: Most hand gesture recognition methods for sign language translation are using vision-based system, through the utilization of single or multiple cameras. The problem is that the user must be located in front of the camera for the system to recognize any hand motions. Furthermore, precision of image processing techniques using cameras are susceptible to illumination of background. Hence, efficient approach by utilizing glove-based system that consists of inertial or orientation sensors are the best method for hand gesture recognition. The purpose of this work is to present a hand gesture recognition method using Artificial Neutral Network that translates gestures based on measurement data from muscle and inertial sensors for a wearable sign language translation device.
