mukham 1.0.4

A basic library to crop largest face from the images.

mukham

Mukham (ముఖం; mook-hum) means face in Telugu! mukham is a simple library to detect the largest face in a given image. The face detection is performed using the DNN algorithm provided by opencv-python. The largest face is the box with largest enclosing area from the bounds of candidate faces detected.

Requirements

This package was written and tested in python 3.7+. This package requires the following dependencies:

    numpy>=1.18.1
    opencv-python>=4.1.2.30

Installation

The package can be installed as follows:

    pip install mukham

Usage

The package can be used through one of the following ways.

Command-line interface

    python -m mukham -i/--input path/to/input_image -o/--ouput path/to/output_largest_face -c/--conf confidence_threshold

This will print the top left and bottom right corner of the bounding box for the largest face. The (i)nput argument is required. The (o)utput path is optional, if given will save the largest face image to this path. The (c)onf is optional, otherwise must be a floating point number between 0 and 1 with default value of 0.8.

Code

    from mukham.detector import detect_largest_face

    bounding_box = detect_largest_face(path_to_input_img, out_path=path_to_output_img, min_conf=confidence_threshold)

The in_path is a required argument for path to the input image. The out_path is an optional keyword argument default to None, if given will save the largest face image to this path. min_conf is also a keyword argument with default set to 0.8. detect_largest_face will return a 2x2 array of the top left and bottom right corner of the bounding box for the largest face.

Tests

Unit tests have also been written which can be run as follows:

    cd path/to/mukham_package
    python -m unittest tests/test.py