Morse Gesture Input — Hibiki Shimizu Works

Full write-up coming soon — Detailed documentation for this project is currently being prepared.

Overview

This HCI (Human-Computer Interaction) project explores an alternative input method for users with physical constraints or in situations where a keyboard is impractical. By holding the hand up to a webcam and making simple gestures, the user enters dots and dashes that are decoded into alphanumeric characters in real time. The system runs entirely on a standard laptop CPU with no specialised hardware.

Key Technologies

Hand tracking: Google MediaPipe Hands — detects 21 landmarks per hand at real-time speed
Gesture classification: Rule-based classifier maps finger-bend states (fist / open / point) to "dot," "dash," and "space"
Morse decoding: Lookup table converts dot/dash sequences to letters and digits
UI: Real-time video overlay via OpenCV shows current input state and decoded text
Stack: Python / MediaPipe / OpenCV

Status

Working prototype stage. A demo video and link to the code repository will be added to this page shortly.