⚽️🥅🤾🏼‍♂️ Football Video Analysis

Overview

Football match video analysis tool using AI and computer vision. Tracks players, analyzes team tactics, and generates performance insights. Built with Python, YOLOv8, and Flask.

🚀 Key Features

Intelligent Object Detection

• Player & Ball & Pitch Tracking: Precise detection using Roboflow models
• DeepSORT Tracking: Consistent object identification across video frames
• Team Classification: Intelligent player team segregation using SigLIP - Sigmoid Loss for Language Image Pretraining

Advanced Analytics

• Player Speed Calculation: Real-time movement velocity estimation
• 2D Pitch Projection: Intuitive field representation
• Voronoi Diagram Generation: Visualize player control zones

Technical Stack

• Deep Learning: Hugging Face, Ultralytics YOLOv8
• Web Framework: Flask
• Object Detection: Roboflow
• Classification: SigLIP

🛠 Quick Setup

Prerequisites

• Python 3.11.0
• GPU recommended for optimal performance

Installation Steps

# Clone Repository
git clone https://github.com/AkbarSheikh-debug/FootBall_AI.git
cd FootBall_AI

# Create Virtual Environment
conda create -p venv python==3.11.0 -y
conda activate venv/ # Activate the virtual environment

# Install Dependencies
pip install -r requirements.txt

🎥 Usage

1. Start the Flask server:

   python -u server.py

2. Access web interface: http://127.0.0.1:5000/

3. Upload football match video (.mp4 or .avi)

4. Process and analyze video

📂 Project Structure

Directory/File	Purpose
main.py	Core video processing logic
server.py	Web application backend
functions/	Annotation and processing utilities
sports/	Sports-specific modules
static/	Web assets and processed videos

💡 Performance Tips

• Use high-quality input videos
• Prefer GPU-enabled machines
• Expect longer processing times for extended videos