Webcam-Based Real-Time Face Filters via Facial Landmark Detection

demo.mp4

Table of Contents

• Table of Contents
• Intruduction
• Features
• Project Structure
• Getting Started
  • Requirements
  • Installation
  • Usage
• Contributing
• License

Intruduction

This application applies dynamic filters to your face using a webcam in real time. It leverages facial landmark detection and offers several filters, including facial landmark overlays, blur, sunglasses, and mustache, with live updates based on face orientation.

Features

1. Real-Time Face Filters: Transform your face live using your webcam. The application processes video frames in real time, applying various filters that adapt to your face's movements and orientation.
2. Facial Landmark Detection: Utilizes MediaPipe's Face Mesh to detect and overlay facial landmarks accurately. This feature provides a detailed map of facial features, allowing other filters to align perfectly with the face.
3. Dynamic Filters:
   • Blur Filter: Apply a Gaussian blur to the face region, which softens facial features. The blur intensity adjusts to cover the entire face, ensuring a smooth and consistent effect.
   • Sunglasses Filter: Adds a pair of sunglasses to the face that adjusts automatically based on the face's orientation. The sunglasses image resizes and rotates in real-time to fit and align with the eyes, enhancing realism and comfort.
   • Mustache Filter: Overlays a mustache on the face that adjusts in size and position according to facial landmarks. The mustache’s placement dynamically adapts to ensure it stays correctly positioned just below the nose, providing a natural appearance.
4. Responsive Filter Application: Filters are applied in real-time, responding to changes in facial expressions and movements. The application maintains high performance and responsiveness, ensuring a smooth user experience.

Project Structure

The project follows a specific structure to organize its files and directories:

├── assets/
│   ├── sunglasses.png                 # Image file for the sunglasses filter.
│   └── mustache.png                   # Image file for the mustache filter.
│
├── src/
│   ├── face_filters.py                # Module containing filter application functions.
│   ├── facial_landmark_detection.py   # Module for detecting and drawing facial landmarks.
│   ├── webcam_capture.py              # Module for capturing video from the webcam and applying filters.
│   ├── webcam_constants.py            # File for storing constant values related to webcam and filters.
│   ├── main.py                        # Main script for initializing the webcam and filter application.
│   └── menu.py                        # Module for displaying the on-screen menu.
│
├── requirements.txt                   # Lists the project's dependencies.
├── .gitignore                         # Specifies which files and directories should be ignored by Git version control.
└── README.md                          # Documentation file providing information about the project.

Getting Started

Requirements

• Python 3.6 or higher
• OpenCV
• MediaPipe
• NumPy

Installation

1. Install Python: Ensure Python 3.X is installed. If not, download and install it from python.org.
2. Clone the Repository:

   git clone https://github.com/Roodaki/RealTime-Webcam-Face-Filters
   cd RealTime-Webcam-Face-Filters/
   

3. Install required packages: Use pip to install the necessary Python libraries:

   pip install -r requirements.txt
   

Usage

Run the main script to start the application:

python main.py

Use the following keys to switch between filters:

• 0: No Filter
• 1: Facial Landmark Filter
• 2: Blur Filter
• 3: Sunglasses Filter
• 4: Mustache Filter
• q: Quit the application

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

For any questions or feedback, please contact amirhossein.rdk@gmail.com.

License

This project is licensed under the MIT License - see the LICENSE file for details.