Copyright © 2025 webdevpathiraja All rights reserved. This project is licensed under the MIT License.

🫲🏼👌🏼Real-Time Hand Tracking with MediaPipe 🤚🏼👋🏼

This project uses MediaPipe and OpenCV to track hands in real-time using a webcam. It detects hand landmarks and draws them on the frame, displaying the tracking in a live video window. The hand landmarks are displayed as red dots, and the connections between the landmarks are drawn with green lines.

Prerequisites

To run this project, make sure you have the following Python libraries installed:

• opencv-python (for webcam capture and image processing)
• mediapipe (for hand tracking)

You can install these libraries using pip:

pip install opencv-python mediapipe

How to Run the Project 🦾🔥

1. Clone this repository to your local machine:

   git clone https://github.com/webdevpathiraja/AI-Hand-Tracking-on-CPU-in-Real-Time.git
   cd AI-Hand-Tracking-on-CPU-in-Real-Time

2. Install the necessary dependencies:

   pip install -r requirements.txt

3. Run the script:

   python main.py

   This will open a webcam window showing the tracked hand landmarks. The landmarks will be drawn as red dots, and green lines will represent the connections between the landmarks.

4. Press q to exit the program.

How It Works 👷🏻‍♀️🪜

• MediaPipe Hands is used to detect the hand landmarks and their connections in real-time. Each hand's landmarks are stored in a list of 21 points, each with 3D coordinates (x, y, z).
• The script starts by opening the webcam feed using OpenCV. It then processes each frame to detect hands.
• Once hands are detected, the landmarks and connections are drawn on the frame using OpenCV.
• The frame is displayed in a window, and you can press q to exit the application.

Landmarks Information ⭕️❎

Each hand has 21 landmarks, numbered from 0 to 20. Here are the main landmarks:

• 0: Wrist
• 1-4: Thumb (with points for the thumb base, tip, etc.)
• 5-8: Index finger (base to tip)
• 9-12: Middle finger (base to tip)
• 13-16: Ring finger (base to tip)
• 17-20: Little finger (base to tip)

These landmarks are used to draw connections between each joint in the hand.