🚗 Vehicle Detection And Counting using OpenCV

This project is a vehicle counting system built with Python and OpenCV. It detects and counts vehicles in real-time from a video file or webcam feed using background subtraction and object tracking techniques.

🧠 Features

• Real-time vehicle detection and tracking
• Automatic up/down direction counting
• Works with both video files and live webcam feed
• Background subtraction using MOG/MOG2
• Noise reduction with morphological operations
• Adjustable parameters (count line position, object size, etc.)
• Displays total, up, and down counts in real-time

🧩 Requirements

• Python 3.8 or later
• OpenCV
• NumPy

Install dependencies with:

pip install opencv-python numpy

🚀 Usage

From a video file

python vehicle_counter.py --video video.mp4 --count-line-position 300

From a webcam

python vehicle_counter.py --camera

⚙️ Arguments

Argument	Description	Default
--video	Path to video file	video.mp4
--camera	Use webcam instead of video file	False
--device	Camera device index	0
--count-line-position	Y position of counting line	2/3 of frame height
--min-width-react	Minimum bounding box width	80
--min-height-react	Minimum bounding box height	80

📊 Output Example

The system displays the processed video with bounding boxes, a counting line, and live vehicle counts:

• Blue → Vehicles moving up
• Orange → Vehicles moving down
• Magenta → Default detection state

Example output display:

Up: 3
Down: 5
Total: 8

🏁 Project Structure

├── vehicle_counter.py
├── video.mp4
├── screenshot.png
└── README.md

💡 Author

👤 Shariful
🔗 GitHub Profile

🏷️ Tags

python opencv computer-vision vehicle-detection object-tracking traffic-analysis background-subtraction video-processing ai image-processing

---

📸 Repository: pythonicshariful/-Vehicle-Detection-And-Counting-using-OpenCV