mediapipe_face_align

mediapipe_face_align is a Python package for aligning faces in images using the MediaPipe Face Mesh model and transforming facial landmarks.

Installation

You can install the package using pip:

pip install mediapipe_face_align

Usage

Import the package and use the process function to align faces and get transformed landmarks.

import mediapipe_face_align

# Load an image
image_path = "sample.png"

# Process the image and get the aligned face and landmarks
aligned_face, transformed_landmarks = mediapipe_face_align.process(image_path)

# Display the aligned face or use the data as needed
# You can use OpenCV to display the image:
# cv2.imshow('Aligned Face', aligned_face)

Example

Here's an example of how to use the mediapipe_face_align package to align faces in an image:

import cv2
import mediapipe_face_align

# Load an image
image_path = "sample.png"

# Process the image and get the aligned face and landmarks
aligned_face, transformed_landmarks = mediapipe_face_align.process(image_path)

# Display the aligned face
cv2.imshow('Aligned Face', aligned_face)
cv2.waitKey(0)
cv2.destroyAllWindows()

Features

Aligns faces in images using facial landmarks.
Provides transformed landmarks for further analysis or processing.
Easy integration with OpenCV and other image processing libraries.

Documentation

For more information and detailed documentation, please visit the GitHub repository.

License

This package is distributed under the MIT License. See the LICENSE file for details.