Study on Hand Gesture Recognition using a Wearable Sensing Glove and Forearm Surface-Mounted EMG

Abstract

There are many techniques for hand recognition. Each suited for certain applications due to their form factor and needs. The most popular is vision-based system by using either single or multiple cameras. Not to forget, Kinect and Leap Motion Controller also part of vision-based system. The problem of the architecture is the user must be in front of the camera in order for the system to recognize any hand motions which limit the portability of the system. Furthermore, precision of image processing techniques using cameras are susceptible to illumination of background. The aim of this study is to document the potential of applying sEMG at forearm together with the effectiveness of sensory glove as the input to a hand gesture system. The sEMG will record the information produce during forearm muscle action. The flex sensors attached on the glove will produce the biofeedback of the fingers by tracking fine movement of finger joints. This will monitor and measure the subject’s Flexor Digitorum Superficialis signal when performing hand movements. The subject need to wear a glove with flex sensors fixed at every finger which allow simultaneous measurement of the flexion and extension force.  Both systems are coupled with a microcontroller that transmit the information gathered from activities to a computer for recording and analysis purposes. Hand gestures are widely applied on different type of applications, ranging from human safety, such as warning, controlling, and directing (robot/human), to pleasure, such as sports and games. In addition, hand gesture system is vastly used as a communication tool between hearing impaired person and normal person, thru the usage of sign language. This study will help to develop better hand gesture system that leads to efficient application especially in reducing the communication gap between the deaf community and normal people without neglecting the system’s portability.