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.
- ✅ 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
- person
- car
- dog
- cat
- truck
- bicycle
- motorcycle
- bus
- bottle
(based on COCO dataset classes)
- Python
- OpenCV
- Ultralytics YOLOv8
- Deep SORT tracking
- Streamlit
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
- Clone this repository
git clone https://github.com/aryansengar007/Object-Detection-and-Tracking.git
cd Object-Detection-and-Tracking- Install dependencies
pip install -r requirements.txt-
Download YOLOv8 model weights (e.g.
yolov8n.pt)
From: https://github.com/ultralytics/ultralytics -
If using virtual environment, activate it
venv\Scripts\activate- Install necessary libraries
pip install ultralytics opencv-python deep_sort_realtime streamlit- Run the app
streamlit run main.py
Aryan Sengar
💻 Passionate about AI, Computer Vision, and Intelligent Systems
📍 Gurgaon, India
💬 Feel free to drop a ⭐ if you find this helpful!