Evaluation of leap motion controller usability in development of hand gesture recognition for hemiplegia patients

Abstract

A hand gesture recognition system is developed for hemiplegia patients to undergo rehabilitation which can encourage patients’ motor function. The Leap Motion controller has been studied to detect human hand motion for development of hand gesture controlled robotic arms. It was shown that the Leap Motion sensor is useful to obtain the coordinate position and orientation of each human finger, palm and wrist movements. A set of test program has been designed using healthy hand to investigate the accuracy and reliability of the sensor. The test results show the effectiveness of the device in the recognition of the human hand gestures with a high accuracy rate of 100% for opening and closing of hand, 97.61% for whole hand tapping and 99.6% for right movement while 98.71% for left movement of whole hand lateral rotation.