Real-Time Sign Language to Text Editor ✋📝
This project is a computer vision application that translates hand gestures into text in real-time. It uses OpenCV for video processing and MediaPipe for hand landmark detection.
🌟 Features
-
Sign Detection: Recognizes 5 alphabet signs: A, B, L, V, Y.
-
Smart Typing: Hold a sign for 1.5 seconds to type the character.
-
Space & Delete: * Show an Open Hand (5 fingers) to add a Space.
-
Show a Fist (0 fingers) to Delete the last character.
-
Visual Interface: Shows a live loading bar and the current sentence on screen.
🛠️ Requirements
-Python 3.x
-OpenCV
-MediaPipe
🚀 How to Run (The Easy Way)
I have created a one-click launcher to make this easy!
1 .Download this repository.
-
Double-click the run.bat file.
-
🪄 It will automatically install all required libraries.
-
🎥 It will launch the camera application instantly.
-
🎮 How to Use
-
Show a sign to the camera.
-
Hold it until the green loading bar fills up.
-
The letter will be added to your sentence!
-
Press 'q' to quit the application.
🛠️ How it Works We will check the "state" of each finger (Up or Down).
-
Sign 'A': All fingers folded, Thumb is Up (or tucked).
-
Sign 'B': All 4 fingers Up, Thumb tucked in.
-
Sign 'L': Thumb and Index Finger Up.
-
Sign 'V': Index and Middle Finger Up.
-
Sign 'Y': Thumb and Pinky Finger Up.
🧠 Tech Stack
-
OpenCV: For video capture and image processing.
-
MediaPipe: For extracting 21 hand landmarks.
-
Python: Core logic using geometric rules (Cartesian coordinates).