A 3D Gesture Modeling Method and Its Application in 3D Gesture Prediction

Abstract

In three-dimensional gesture recognition, it has always been a difficult problem to determine the accurate three-dimensional coordinates of the real hand. Direct use of image recognition cannot obtain complete information of hand joints. Even with the help of external equipment (such as data gloves and infrared recognition), there will be errors, which affect the accuracy of gesture recognition. However, on the other hand, the establishment of the virtual 3D model of the hand avoids this limitation and provides a new scheme for the 3D coordinate acquisition of the real hand. The research results of this paper have application value in sign language recognition, virtual environment interaction, user interface, robot arm grabbing, etc. In the field of sign language recognition, the recognized sign language can be translated to help deaf-mutes communicate with most people in sign language. In the interaction of the virtual environment, gestures in the real environment can be used to control the virtual hands in the virtual environment to make corresponding actions. In the application of user interface, users can communicate instructions to computers through gestures. In the field of grasping by robot manipulators, gestures can be used to control manipulators to complete tasks.