🎯 Object Detection and Tracking (YOLOv8 + Deep SORT)

This project is a real-time object detection and tracking system built using the Ultralytics YOLOv8 models and Deep SORT tracking. It includes a clean and interactive Streamlit interface with webcam, video, and image input support.

🚀 Features

• ✅ YOLOv8 Model Variant Selection (n, s, m, l, x)
• ✅ Webcam, video file, and image input support
• ✅ Deep SORT tracking (toggleable)
• ✅ Real-time FPS counter
• ✅ Class-wise object count summary
• ✅ Filter detection by object classes
• ✅ Confidence threshold control
• ✅ Detection log console
• ✅ Download processed video and CSV summary
• ✅ Fully interactive and professional Streamlit GUI

📸 Sample Classes Supported

• person
• car
• dog
• cat
• truck
• bicycle
• motorcycle
• bus
• bottle

(based on COCO dataset classes)

🛠️ Technologies Used

• Python
• OpenCV
• Ultralytics YOLOv8
• Deep SORT tracking
• Streamlit

📂 Project Structure

Object-Detection-and-Tracking/
│
├──venv/
├── main.py                  # Streamlit GUI logic
├── detect_and_track.py      # Core detection/tracking logic
├── utils.py                 # Model descriptions and class list

▶️ How to Run

1. Clone this repository

git clone https://github.com/aryansengar007/Object-Detection-and-Tracking.git
cd Object-Detection-and-Tracking

2. Install dependencies

pip install -r requirements.txt

3. Download YOLOv8 model weights (e.g. yolov8n.pt)
   From: https://github.com/ultralytics/ultralytics

4. If using virtual environment, activate it

venv\Scripts\activate

5. Install necessary libraries

pip install ultralytics opencv-python deep_sort_realtime streamlit

6. Run the app

streamlit run main.py

🖥️ Demo Screenshots

Dashboard

Result

Generated File and Video

🙋‍♂️ Developed By

Aryan Sengar
💻 Passionate about AI, Computer Vision, and Intelligent Systems
📍 Gurgaon, India

© 2025 Aryan Sengar – All Rights Reserved
Unauthorized copying is strictly prohibited.

💬 Feel free to drop a ⭐ if you find this helpful!