3D Gesture-based Interaction Interface Oriented to Mental Model Expression

摘要

The process of the three-dimensional interaction has different motion style for different users. To reflect the users' different motion style, this paper presents a gesture style expression algorithm based on monocular vision from the three aspects of speed, trajectory and perspective. Firstly, we establish the gesture behavior model under the fundamental experimental platform using camera, data glove and position tracker. Secondly, we put emphasis on building the speed, trajectory and perspective models according to the behavior model. Finally, we get the styles of gesture based on motion models and complete the interactive task. The main contribution of this paper is that we establish mathematical relation between the real hand and gesture animation for users from the perspective of gesture behavioral model. Our experimental results demonstrate that the proposed approach fully reflects the different styles for users. The overall average tracking accuracy error is 12% compared with data glove in different speed. At the same time, the system based on our algorithm can reduce the cognitive load for users and provide an enhanced user experience.