3d tennis swing tracker

pose detection of tennis swing paths from video footage using MediaPipe

• extracts joint coords, computes geometric transformation to normalize 3d data (uses court boundaries + net as camera angle baselines)
• cross-video comparative analysis across database of professional 3d swing paths
• explores a "similarity scoring" to suggest optimal playing style

demo: https://youtu.be/HbyJnkKpvEI

Setup

1. Create a virtual environment (recommended):

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

2. Install dependencies:

pip install -r requirements.txt

3. Create a database folder and add professional tennis videos for comparison.

Usage

1. Run the main script:

python main.py

2. Follow the prompts to:
   • Select a video file for analysis
   • Choose comparison parameters
   • View the analysis results

Project Structure

• main.py: Main script for video processing and analysis
• pose_estimation.py: Pose estimation and tracking module
• movement_analysis.py: Movement analysis and comparison module
• database/: Directory for storing professional tennis videos
• output/: Directory for storing analysis results