A Gesture Design Toolkit for Computer Games

摘要

Recently, vision based hand gesture interfaces have become very common. However, the development of such applications is very difficult. It requires sophisticated image processing and machine learning knowledge that is out of reach for many traditional game developers. They need an auxiliary tool to design vision based interfaces. This paper describes the motivation, design and development of a toolkit for vision based interactive games. It has the following characteristics: a flexible interface to facilitate adding new classifiers; a dynamic configuration management mechanism for all classifiers; and a visual user interface to define high-level semantic gesture. Evaluation results show that it can provide a unified platform and a general solution for vision based computer games.