The practical adaption of interface solutions for visual impaired and blind people is limited by simplicity and usability in practical scenarios. Different solutions (e.g. Drishticite{helal2001drishti}) focuses upon speech or keyboard interfaces, which are not efficient or transparent in every-day environments. As an easy and practical way to achieve human-computer-interaction, in this paper hand gesture recognition was used to facilitate the reduction of hardware components. Additionally a qualitative user study was performed to compare learning curves of different subjects with and without prior knowledge of gesture recognition devices, interpreting the readings from a sensitive surface by machine learning algorithms. The user study was made using well-known machine learning algorithms applied to recognizing symbols from the graffiti handwriting system [2] and the WEKA data mining software [3] for comparing individual machine learning approaches.
展开▼
