Vehicle counting provides critical information to traffic analysis problems (like monitoring flow, peaks/jams, etc.) and other related traffic issues like highway management. Having an good and accurate understanding of traffic flow can help for urban planners to be able to make more informed decision to improve the general quality of life for the citizens. While there already exist numerous methods to accomplishing the task already, from manually counting to special censors (and many other alternatives), developing smart systems to automatically count in real time can be incredibly useful and save much time and effort.

python predict.py model='name_of_your_model.pt' source='name_of_your_input_file.mp4' show=True

https://colab.research.google.com/github/roboflow-ai/notebooks/blob/main/notebooks/train-yolov8-object-detection-on-custom-dataset.ipynb

pip install -e '.[dev]'

cd ultralytics/yolo/v8/detect

https://github.com/nwojke/deep_sort

https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8l.pt