Cropped faces after detection in an image using face_detection module and save them.

face_detection_n_crop.py

import face_recognition
from PIL import Image

np_image = face_recognition.load_image_file("avengers2.jpg")

# face_locations returns a list of locations for each face 
# detected in the picture, each list item contains a tuple 
# of 4 items in the format of (top, right, bottom, left)
face_locations = face_recognition.face_locations(np_image)

image_obj = Image.open("avengers2.jpg")
seq = 0

for face_location in face_locations:
    # Be careful in ordering top, right, bottom, left
    top, right, bottom, left = face_location
    cropped_image = image_obj.crop((left, top, right, bottom))
    
    cropped_image.save('avengers2_cropped_'+str(seq)+'.jpg')
    cropped_image.show()
    seq += 1

The face recognition module: https://github.com/ageitgey/face_recognition