This project demonstrates hand gesture recognition using the Mediapipe library. It captures hand gesture points from a video, visualizes them in real-time, and classifies gestures based on pre-recorded templates.
- Hand Gesture Capture: Utilizes the Mediapipe library to capture hand gesture points.
- Real-time Visualization: Visualizes hand gestures in real-time during video playback.
- Template Recording: Records and creates templates for specific gestures (e.g., "Hello" and "No Hello").
- Gesture Recognition: Uses a gesture recognizer to classify gestures in a new video.
-
Install dependencies:
pip install mediapipe opencv-python matplotlib dollarpy
-
Run the
gesture_recognition.py
script to capture and recognize hand gestures.python gesture_recognition.py
Ensure that your camera is properly configured or update the video paths in the script. Customize the template recording section for your specific gestures. Requirements Python 3.x mediapipe opencv-python matplotlib dollarpy License This project is licensed under the MIT License.
Feel free to contribute or report issues!