Dexterous Hand Motion Classification and Recognition Based on Multimodal Sensing

Human hand motions analysis is an essential research topic in recent applications, especially for dexterous robot hand manipulation learning from human hand skills. It provides important information about gestures, moving, speed and the control force captured via multimodal sensing technologies. This paper presents a comprehensive discussion of the nature of human hand motions in terms of simple motions, such as grasps and gestures, and complex motions, e.g. in-hand manipulations and re-grasps. And then, a novel multimodal sensing based hand motion capture system is proposed to acquire the sensory information. By using an adaptive directed acyclic graph algorithm, the experimental results show the proposed system has a higher recognition rate compared with those with individual sensing technologies.