Face Recognition With OpenCV Python

In this tutorial we are going to learn how to perform Facial recognition with OpenCV python in Pycharm. Head on to our Pycharm project. Here we will install the required packages. Below is the list.

• cmake
• dlib (try installing lower version)
• face_recognition
• numpy
• opencv-python

Importing

First we will import the relevant libraries

import face_recognition

import cv2

import numpy as np

Loading Images and Converting to RGB.

The Face Recognition package consists of a load image function that loads the image. Once imported the image has to be converted to RGB.

imgmain = face_recognition.load_image_file('ImageBasics/Bryan_Cranst.jpg')
imgmain = cv2.cvtColor(imgmain, cv2.COLOR_BGR2RGB)
imgTest = face_recognition.load_image_file('ImageBasics/bryan-cranston-el-camino-aaron-paul-1a.jpg')
imgTest = cv2.cvtColor(imgTest, cv2.COLOR_BGR2RGB)

For this you can use any persons photo and add it into the directory.

Find Faces Locations and Encodings

In the second step we will use the true functionality of the face recognition library. First we will find the faces in our images . This is done using HOG (Histogram of Oriented Gradients) at the backend. Once we have the face they are warped to remove unwanted rotations. Then the image is feed to a pretrained neural network that out puts 128 measurements that are unique to that particular face. The parts that the model measures is not known as this is what the model learns by itself when it was trained. Lucky for us all this is done is just 2 lines of code. Once we have the face locations and the encodings we can draw rectangles around our faces.

faceLoc = face_recognition.face_locations(imgmain)[0]
encodeElon = face_recognition.face_encodings(imgmain)[0]
cv2.rectangle(imgmain, (faceLoc[3], faceLoc[0]), (faceLoc[1], faceLoc[2]), (255, 0, 255), 2)

faceLocTest = face_recognition.face_locations(imgTest)[0]
encodeTest = face_recognition.face_encodings(imgTest)[0]
cv2.rectangle(imgTest, (faceLocTest[3], faceLocTest[0]), (faceLocTest[1], faceLocTest[2]), (255, 0, 255), 2)

Compare Faces and Find Distance

Once we have the encodings for both faces, then we can compare these 128 measurements of these two faces to find similarities. To compare the package uses one of the most common machine Learning methods linear SVM classifier. We can use the compare_faces function to find if the faces match. This function returns True or False. Similarly we can use the face_distance function to find how likely is the faces match in terms of numbers. This is helpul particularly when there are multiple faces to detect from.

results = face_recognition.compare_faces([encodeElon], encodeTest)
faceDis = face_recognition.face_distance([encodeElon], encodeTest)
print(results, faceDis)
cv2.putText(imgTest, f'{results} {round(faceDis[0], 2)}', (50, 50), cv2.FONT_HERSHEY_COMPLEX, 1, (0, 0, 255), 2)

Full Code

import cv2
import face_recognition

imgmain = face_recognition.load_image_file('ImageBasics/Bryan_Cranst.jpg')
imgmain = cv2.cvtColor(imgmain, cv2.COLOR_BGR2RGB)
imgTest = face_recognition.load_image_file('ImageBasics/bryan-cranston-el-camino-aaron-paul-1a.jpg')
imgTest = cv2.cvtColor(imgTest, cv2.COLOR_BGR2RGB)

faceLoc = face_recognition.face_locations(imgmain)[0]
encodeElon = face_recognition.face_encodings(imgmain)[0]
cv2.rectangle(imgmain, (faceLoc[3], faceLoc[0]), (faceLoc[1], faceLoc[2]), (255, 0, 255), 2)

faceLocTest = face_recognition.face_locations(imgTest)[0]
encodeTest = face_recognition.face_encodings(imgTest)[0]
cv2.rectangle(imgTest, (faceLocTest[3], faceLocTest[0]), (faceLocTest[1], faceLocTest[2]), (255, 0, 255), 2)

results = face_recognition.compare_faces([encodeElon], encodeTest)
faceDis = face_recognition.face_distance([encodeElon], encodeTest)
print(results, faceDis)
cv2.putText(imgTest, f'{results} {round(faceDis[0], 2)}', (50, 50), cv2.FONT_HERSHEY_COMPLEX, 1, (0, 0, 255), 2)

cv2.imshow('Main Image', imgmain)
cv2.imshow('Test Image', imgTest)
cv2.waitKey(0)