This repository contains Python code to count the number of cars, trucks, buses, and motorbikes in a video using YOLOv8 object detection. It also tracks the detected vehicles and displays the count in real-time.
Car Counter 
Requirements Before running the code, make sure you have the following Python modules installed:
numpy opencv-python cvzone ultralytics sort You can install these modules using pip with the following commands:
pip install numpy opencv-python cvzone ultralytics pip install git+https://github.com/abewley/sort.git
Clone the repository:
git clone https://github.com/biswadeep-roy/cars-counter-yolov8.git
Download the YOLOv8 weights file (e.g., yolov8l.pt) and place it in the root directory of the repository.
Create a mask image for region of interest (ROI) using a tool like Canva. Save the mask as mask.png and place it in the root directory of the repository.
Run the code:
python Car_Counter.py
The code will open a video stream and display the vehicles detected in real-time. It will also count the number of vehicles that cross a predefined line in the video.
You can modify the classNames list to include or exclude specific object classes for detection. Adjust the confidence threshold (conf > 0.3) to control the detection sensitivity. Modify the limits list to change the position of the counting line in the video.
This project is licensed under the MIT License - see the LICENSE file for details..
YOLOv8: Ultralytics YOLOv5 GitHub Repository SORT (Simple Online and Realtime Tracking): SORT GitHub Repository