Prosthetic Hand Control — EMG (2D)

This 2D companion runs the exact same signal chain as the 3D version — a per-grasp motor-unit firing model synthesising multi-channel EMG, RMS/MAV/ZC feature extraction over a 256-sample sliding window, and a genuinely fitted pooled-covariance LDA classifier — through a plain canvas view built for reading the pipeline rather than orbiting a scene: live scrolling channel traces show which electrodes the chosen grasp is actually driving, a feature-space scatter plots the fitted class clusters against your current sample, a schematic hand's finger curl is wired directly to the classifier's live prediction, and a held-out accuracy readout reports how well the classifier actually separates the six grasps at your current noise and channel-count settings.