Hand Gesture Recognition for Game 3D Object Using The Leap Motion Controller with Backpropagation Method

Abstract

Computer games continue to grow and are used by people and become a research topic in the field of computer vision. Leap Motion Controller is a computer vision technology that is able to read human movements quickly. In this research, it is moving 3D animation using hand gestures with the help of Leap Motion Controller. The input of hand motion data that emits from Leap Motion is analyzed using the backpropagation method. This artificial neural network pattern uses three input layer network patterns, four hidden layers, one output layer. The data obtained are cultural and hand index data. Pointable and hand are part of finger tracks issued by the Leap Motion sensor. The type of movement used to move 3D objects in this research is a swipe to wave, circle to go, Keytap to walk, Screencap to advance or run. The data needed in the design of backpropagation artificial neural applications is to take variables from the data obtained from Pointables and hands to the coordinates of the x, y, and z-axes. The resulting accuracy result is 96.7%. In addition, backpropagation output to control 3D animation.