Zynq SoC-based Hand Motion Recognition System Using EMG Signals

People who have lost a hand or upper limb face significant challenges in mobility and independence, but active prostheses provide a pathway to restoring autonomy. This work presents a hardware implementation on a Zynq System-on-Chip (SoC) to recognize three distinct hand movements using electromyography (EMG) signals and a support vector machine (SVM) classifier. The software implementation achieved a recognition accuracy of 98.79%, showcasing the system's robustness. In the hardware implementation, the architecture was further optimized using parallelism techniques, which significantly enhanced processing efficiency. This optimization resulted in a sixfold improvement in speed compared to the baseline non-optimized design. These advancements pave the way for real-time, low-latency solutions in prosthetic control, enhancing user experience and functionality