Object Detection and Tracking System

This project combines YOLO (You Only Look Once) object detection with DeepSORT tracking to create a comprehensive system for real-time object detection and tracking across multiple input sources. The system supports video files, webcam feeds, RTSP streams, and static images.

🌟 Features

• Multiple Input Sources Support:

  • Video file processing
  • Webcam real-time detection
  • IP Camera (RTSP) streaming
  • Static image detection

• Advanced Tracking:

  • Object tracking with unique IDs
  • Track duration monitoring
  • ID switch detection
  • Consistent color coding per track

• Visualization:

  • Semi-transparent bounding boxes
  • Real-time FPS counter
  • Object class labels
  • Track duration display
  • Track ID visualization

• Comprehensive Metrics:

  • Processing performance stats
  • Detection confidence analysis
  • Tracking performance metrics
  • Bounding box statistics

• User-Friendly Interface:

  • Interactive menu system
  • Custom weight file selection
  • Operation cancellation (ESC)
  • Return to menu option
  • Progress bar for video processing

🛠️ Installation

1. Clone the repository:

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

2. Create and activate virtual environment (optional but recommended):

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

3. Install required packages:

pip install -r requirements.txt

📦 Requirements

• Python 3.8+
• OpenCV
• Ultralytics YOLO
• Deep SORT Real-Time
• NumPy
• Pandas
• tqdm

🚀 Usage

1. Run the main script:

python main.py

2. Select from available options:

=== Object Tracking System ===
1. Process Video File
2. Use Webcam
3. Use IP Camera (RTSP)
4. Process Image
5. Exit

3. Enter the path to your YOLO weights file (press Enter for default 'best.pt')

4. Provide required input paths based on your selection

Controls During Operation:

• Press 'q' to stop processing
• Press 'ESC' to cancel operation
• After processing:
  • Press 'M' to return to main menu
  • Press 'Q' to quit program

📊 Output Metrics

The system provides comprehensive metrics including:

• Processing Summary:

  • Processed Frames
  • Time Taken
  • Average FPS
  • Video Duration

• Detection and Tracking Metrics:

  • Total Detections
  • Unique Tracks
  • Objects per Frame
  • ID Switches

• Confidence Metrics:

  • Average Confidence
  • Min/Max Confidence

• Tracking Performance:

  • Track Duration
  • Track Length
  • Bounding Box Areas

🖼️ Sample Output

📊 Processing Summary:
  • Processed Frames: 300
  • Time Taken: 10.5 seconds
  • Average FPS: 28.57
  • Video Duration: 10.00 seconds

📊 Detection and Tracking Metrics:
  • Total Detections: 450
  • Unique Tracks: 5
  • Average Objects per Frame: 1.50
  • ID Switches: 2

🎯 Use Cases

• Security and Surveillance
• Traffic Monitoring
• People Counting
• Object Movement Analysis
• Retail Analytics
• Sports Analysis

🤝 Contributing

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

👏 Acknowledgments

• YOLOv8 by Ultralytics
• Deep SORT algorithm
• OpenCV team
• Contributor community

👏 Supervisor

• Prof. Chao-Tung Yang, Tunghai University
• Assistant Prof. Ding-Hsiang Huang., Ph.D, Tunghai University
• OpenCV team
• Contributor community

✉️ Contact

Anggi Andriyadi - [email protected] Tunghai University

Project Link: [https://github.com/PamanGie/yolo_deepsort)

Result