Virtual Piano with OpenCV and Mediapipe

Overview

This project implements a virtual piano using OpenCV and Mediapipe, allowing users to play music by interacting with their hands in front of a camera. The application detects hand keypoints using Mediapipe's Hand module and maps them to corresponding piano keys on the screen.

Demo Video

Virtual Piano Demo

final_demo.mp4

Getting Started

Installation

• Clone the repository

git clone https://github.com/NAGAMALLYSRUJAN2329/vir_piano.git
cd vir_piano

• Create new conda environment

conda create -n VirPiano python=3.10

• Activate the environment

conda activate VirPiano

• Install dependencies

pip install -r requirements.txt

Local setup

python main.py

• If you want to change settings of the paino, use the below cli command.

python main.py --model_path "model/hand_landmarker.task" --num_octaves 2 --list_of_octaves "[3,4]" --height_and_width_black "[[5,8],[5,8]]" --shape "(800,600,3)" --tap_threshold 20 --piano_config_threshold 30 --piano_config 1

Controls

• Quit: Press 'Q' key.

For Streamlit Interface

streamlit run streamlit_main.py

Troubleshooting

• If there is no hand tracking, try adjusting the camera settings and lighting conditions and try to move the hand slowly in front of the camera.
• Ensure that your Python environment meets the specified requirements.

Contributers

NAGAMALLYSRUJAN2329
Prachi-0205
Prkhar05
xenom2004

Contributing

Contributions are welcome!