Advanced Interfaces for HMI in Hand Gesture Recognition

The present thesis investigates techniques and technologies for high quality Human Machine Interfaces (HMI) in biomedical applications. Starting from a literature review and considering market SoA in this field, the thesis explores advanced sensor interfaces, wearable computing and machine learning techniques for embedded resource-constrained systems. The research starts from the design and implementation of a real-time control system for a multifinger hand prosthesis based on pattern recognition algorithms. This system is capable to control an artificial hand using a natural gesture interface, considering the challenges related to the trade-off between responsiveness, accuracy and light computation. Furthermore, the thesis addresses the challenges related to the design of a scalable and versatile system for gesture recognition with the integration of a novel sensor interface for wearable medical and consumer application.