This project is a Flask-based web application that performs real-time object detection on uploaded videos using YOLOv8 (Ultralytics).
Users can upload a video, and the app streams the detected objects frame-by-frame in the browser.
- Upload video via web interface
- Real-time object detection using YOLOv8
- Live video streaming with bounding boxes
- Lightweight and easy to deploy
- Uses OpenCV for video processing
- YOLOv8 Nano (
yolov8n.pt) - Pre-trained on the COCO dataset
bash
βββ app.py
βββ README.md
βββ yolov8.pt
bash
git clone (https://github.com/Bilal-73/YOLOv8-Real-Time-Video-Object-Detection-with-Flask.git)
cd your-repo
- Upload a video file
- Video is saved in the uploads/ directory
- Frames are processed using YOLOv8
- Annotated frames are streamed back to the browser
- Large videos may cause slow streaming
- GPU is recommended for better performance
- This demo uses a single global video session
Bilal Imran
- π BS Computer Science (AI Specialization)
- π» AI / Machine Learning & Full-Stack Developer
- π Pakistan
GitHub: https://github.com/Bilal-73