deepinvalue / yolov9-supervision-tracking-counting

Object detection, tracking, and counting with YOLOv9 and Supervision.

Detection, Tracking, and Counting with YOLOv9 and Supervision

This repository contains a Jupyter Notebook that offers an introduction to implementing object detection, tracking, and counting using YOLOv9 and Supervision. Given that YOLOv9 has been released very recently, we are in the exciting early stages of exploring its capabilities and limitations. Please note that as the model and its ecosystem are still being refined, you might encounter some bugs. I aim to keep the code updated with the latest advancements and workarounds.

Full Frame Tracking and Counting

Selective Area Tracking and Counting

Features

• Object detection with YOLOv9
• Real-time object tracking
• Object counting, including selective area counting
• Extending Supervision's Detections for YOLOv9 results
• Video annotation and processing techniques

Acknowledgments

Thanks to everyone involved in YOLOv9's development and the Supervision library, as well as to the broader open-source community for their contributions and support.

Author

Amin Najafgholizadeh Nafouti

Feel free to contact me for any questions or collaborations related to this work.

yolov9-supervision-tracking-counting's Issues

NOT WORKING ON WEBCAM.

I want to see live inference of yolov9 with tracking but When I set,

SOURCE_VIDEO_PATH = 1
TARGET_VIDEO_PATH = "output.mp4"

and ran the Detection and Tracking cell. It does not work and does not show live screen.

获取class id有问题