Cricket Object Detection with YOLOv8

A comprehensive computer vision project for detecting and tracking cricket-related objects in videos using YOLOv8. This system can detect and track:

• Ball - Cricket ball detection and tracking
• Stumps - Wicket stumps identification
• Players - Cricket players on the field
• Umpires - Match officials detection

🚀 Features

• Multi-Object Detection: Simultaneous detection of 4 different cricket objects
• Real-time Tracking: Uses ByteTrack for consistent object tracking across frames
• Video Processing: Full video annotation with bounding boxes and labels
• Statistics: Detailed object count statistics and frame-by-frame analysis
• Easy Annotation: Interactive tool for labeling training data
• Flexible Training: Custom YOLOv8 training pipeline with validation

📁 Project Structure

├── frame_extract.py        # Extract frames from video files
├── label_tool.py           # Interactive annotation tool for labeling frames
├── train_yolov8.py         # Train YOLOv8 model on annotated dataset
├── annotate_video.py       # Process videos with trained model
├── requirements.txt        # Python dependencies
├── cricket_dataset/        # Dataset directory
│   ├── images/            # Training and validation images
│   │   ├── train/         # Training images
│   │   └── val/           # Validation images
│   ├── labels/            # YOLO format annotations
│   │   ├── train/         # Training labels
│   │   └── val/           # Validation labels
│   └── cricket.yaml       # Dataset configuration
└── annotated_videos/       # Output directory for processed videos

🛠️ Installation & Setup

Prerequisites

• Python 3.8 or higher
• CUDA-compatible GPU (recommended for training)

1. Clone the Repository

git clone https://github.com/yourusername/cricket-object-detection-yolov8.git
cd cricket-object-detection-yolov8

2. Create Virtual Environment

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

3. Install Dependencies

pip install -r requirements.txt

4. Download Pre-trained Model (Optional)

The system will automatically download YOLOv8 base models when needed, but you can pre-download:

# This will be done automatically on first run

📖 Usage Guide

Step 1: Extract Frames from Videos

Extract frames from your cricket videos for annotation:

python frame_extract.py

• Place your video files in the project directory
• The script will extract frames at regular intervals
• Frames are saved to cricket_dataset/images/ directory

Step 2: Annotate the Extracted Frames

Use the interactive annotation tool to label objects:

python label_tool.py

Annotation Controls:

• 'n': Next image
• 'p': Previous image
• '0': Select Ball class
• '1': Select Stumps class
• '2': Select Player class
• '3': Select Umpire class
• 's': Save current annotations
• 'c': Clear all annotations for current image
• 'q': Quit annotation tool

How to Annotate:

1. Select a class (0-3)
2. Click and drag to draw bounding boxes around objects
3. Save regularly with 's'
4. Navigate between images with 'n' and 'p'

Step 3: Train the YOLOv8 Model

Train your custom object detection model:

python train_yolov8.py

Training Features:

• Automatic train/validation split
• Progress monitoring with metrics
• Model checkpoints saved automatically
• Best model saved to cricket_detection/training_run/weights/best.pt

Step 4: Process Videos with Trained Model

Annotate new videos using your trained model:

python annotate_video.py

Output Features:

• Bounding boxes around detected objects
• Class labels and confidence scores
• Object tracking IDs
• Frame-by-frame statistics
• Progress monitoring during processing

🎯 Model Performance

The system is designed to detect:

• Ball: Small, fast-moving cricket ball
• Stumps: Wooden wicket posts
• Players: Cricket players in various poses
• Umpires: Match officials on the field

🔧 Configuration

Dataset Configuration

Edit cricket_dataset/cricket.yaml to modify:

• Dataset paths
• Class names
• Training parameters

Model Parameters

Modify training parameters in train_yolov8.py:

• Epochs, batch size, learning rate
• Image size, confidence threshold
• Validation split ratio

📊 Output

Annotated Videos

• Saved to annotated_videos/ directory
• Include bounding boxes, labels, and tracking IDs
• Real-time statistics overlay

Training Metrics

• Model performance metrics
• Validation accuracy
• Loss curves and mAP scores

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/new-feature)
3. Commit your changes (git commit -am 'Add new feature')
4. Push to the branch (git push origin feature/new-feature)
5. Create a Pull Request

👨‍💻 Author

Anuj Dev Singh

• Project Creator & Lead Developer

📋 Requirements

• Python 3.8+
• OpenCV
• PyTorch
• Ultralytics YOLOv8
• Supervision (for tracking)
• NumPy
• PyYAML

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🐛 Troubleshooting

Common Issues:

1. CUDA Out of Memory: Reduce batch size in training
2. Video Not Opening: Check video codec compatibility
3. Model Not Found: Ensure training completed successfully
4. Annotation Tool Issues: Check OpenCV installation

Getting Help:

• Open an issue on GitHub
• Check existing issues for solutions
• Ensure all dependencies are correctly installed

🚀 Future Improvements

• Real-time camera feed processing
• Advanced tracking algorithms
• Mobile app integration
• Web-based annotation interface
• Automated player action recognition
• Match statistics generation

---

Note: This project requires significant computing resources for training. For best results, use a CUDA-compatible GPU.