Chehro is a face detection library for python with a mediapipe backend. It is faster and more accurate comparing to the old HAARCASCADE Classifier way of face detection in python/OpenCV.

First of all, you have to clone this repository to your local machine:

git clone https://github.com/prp-e/chehro

Then, you move to the source directory and run:

pip3 install -e .

NOTE: In some systems(specially Linux based ones) pip3 might be just pip, since most of moderl Linux distributions just left Python2 behind.

In case you want this package to be installed using pip then just run this command:

pip3 install chehro

NOTE: In some systems(specially Linux based ones) pip3 might be just pip, since most of moderl Linux distributions just left Python2 behind.

First, create a new python environment or project. Then you easily can do this:

import cv2
from chehro import Chehro

image = cv2.imread('faces.jpg')
detector = Chehro(image)

result, bboxes = detector.detect()

# In case you might need no bounding box drawn around faces, just run it like this
result, bboxes = detector.detect(draw=False)

NOTE : detect() method returns a tuple made of an image (or the result) and a list of bounding boxed. All bounding boxes are normalized and you don't need to normalize it manually.