Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/harshitwaldia/traffictally

"Detect and count vehicles in videos using OpenCV for accurate analysis in traffic management and monitoring systems."
https://github.com/harshitwaldia/traffictally

computer-vision numpy opencv python3 vehicle-counting vehicle-detection

Last synced: 1 day ago
JSON representation

"Detect and count vehicles in videos using OpenCV for accurate analysis in traffic management and monitoring systems."

Awesome Lists containing this project

README 

        
# 🚗 TrafficTally : Vehicle Detection and Counting Notebook 🚙



This project is a Jupyter notebook application designed to detect and count vehicles in a video stream. It uses OpenCV for image processing and a background subtractor algorithm to identify moving vehicles. The application draws bounding boxes around detected vehicles and counts them as they cross a predefined line in the video frame.



## Features



- **🚘 Vehicle Detection**: Identifies and draws bounding boxes around moving vehicles.

- **📊 Vehicle Counting**: Counts vehicles as they cross a specified line in the video frame.

- **⏱️ Real-Time Processing**: Processes video frames in real-time for live counting.



## Requirements



- 🐍 Python 3.x

- 🖼️ OpenCV

- 🔢 NumPy

- 📓 Jupyter



## Installattion

### Setting Up the Environment



It's recommended to use a virtual environment to manage the dependencies for this project. Follow the steps below to set up the environment using `conda`.



1. Install [Miniconda](https://docs.conda.io/en/latest/miniconda.html) or [Anaconda](https://www.anaconda.com/products/distribution) if you haven't already.



2. Create a new conda environment:



    ```bash

    conda create --name vehicle_counter python=3.8

    ```



3. Activate the environment:



    ```bash

    conda activate vehicle_counter

    ```



4. Install the required packages:



    ```bash

    pip install -r requirements.txt

    ```

## Usage

To run the vehicle detection and counting notebook, follow these steps:



1. Ensure you have a video file named `video3.mp4` in the same directory as your notebook, or modify the code to use a different video file.



2. Launch Jupyter Notebook:

```bash

jupyter notebook

```

3. Open the [`main.ipynb`](main.ipynb) notebook and run all cells to start the vehicle detection and counting process.



4. Run all the cells in the notebook to start the vehicle detection and counting process.



## Explanation of Key Components

- **🔍 Background Subtraction**: Uses cv2.bgsegm.createBackgroundSubtractorMOG() to segment moving objects (vehicles) from the background.


- **✏️ Contour Detection**: Detects contours in the segmented frames and draws bounding boxes around the detected vehicles.


- **🔢 Vehicle Counting**: Defines a counting line and increments the vehicle count whenever a detected vehicle crosses this line.


- **📍 Center Function**: Calculates the center of the bounding box for each detected vehicle.



## Customization

- **Change Count Line Position**:

Modify the count_line_position variable to change the position of the counting line.



- **Adjust Rectangle Size**:

Modify the min_width_rectangle and min_height_rectangle variables to adjust the minimum size of detected vehicles.



- **Use Different Video Source**:

Change the path in cv2.VideoCapture("video3.mp4") to use a different video file or a webcam.



## Code Explanation

The script performs the following steps:



- Imports necessary libraries and initializes video capture.

- Sets parameters for vehicle detection and counting.

- Defines a background subtraction algorithm and contour detection.

- Implements real-time processing to count vehicles crossing a designated line.



## Contributions

Contributions are welcome! If you have any suggestions or improvements, please create a pull request or open an issue on GitHub.