Image_FaceRecognition.py

# -*- coding: utf-8 -*-
"""
Created on Tue Jan 25 14:12:03 2022

@author: Palla
"""

# importing the required libraries
import cv2
import face_recognition

# laoding the image to detect
original_image = cv2.imread('images/testing/harshit.png')

modi_image = face_recognition.load_image_file('images/samples/modi.jpg')
modi_face_encodings = face_recognition.face_encodings(modi_image)[0]

trump_image = face_recognition.load_image_file('images/samples/trump.jpg')
trump_face_encodings = face_recognition.face_encodings(trump_image)[0]

harshit_image = face_recognition.load_image_file('images/samples/harshit.jpg')
harshit_face_encodings = face_recognition.face_encodings(harshit_image)[0]

sanya_image = face_recognition.load_image_file('images/samples/sanya.jpg')
sanya_face_encodings = face_recognition.face_encodings(sanya_image)[0]

known_face_encoding = [modi_face_encodings, trump_face_encodings,harshit_face_encodings,sanya_face_encodings]
known_face_names = ["Narendra Modi", "Donald Trump","Harshit Singh","sanya mehadia"]

image_to_recognize = face_recognition.load_image_file(
    'images/testing/harshit.png')

# select all faces in the image to upsample
all_face_locations = face_recognition.face_locations(
    image_to_recognize, model='hog')

all_face_Encodings = face_recognition.face_encodings(
    image_to_recognize, all_face_locations)

print('There are {} no of faces in this image'.format(len(all_face_locations)))

for current_face_location, current_face_encoding in zip(all_face_locations, all_face_Encodings):
    top_pos, right_pos, bottom_pos, left_pos = current_face_location
    all_matches = face_recognition.compare_faces( 
        known_face_encoding, current_face_encoding)
    name_of_person = 'Unknown'
    if True in all_matches:
        first_match_index = all_matches.index(True)
        name_of_person = known_face_names[first_match_index]

    cv2.rectangle(original_image, (left_pos, top_pos),
                  (right_pos, bottom_pos), (255, 0, 0), 2)
    font = cv2.FONT_HERSHEY_DUPLEX
    cv2.putText(original_image, name_of_person, (left_pos,
                bottom_pos), font, 0.5, (255, 255, 255), 1)
    cv2.imshow("Faces Identified", original_image)