app.py

import face_recognition
import os
from shutil import copy
import cv2
import time
import multiprocessing
import shutil

def capture_image():
    global knownFace
    knownImgPath='known'

    print("Capturing images... ")
    camera = cv2.VideoCapture(0)
    for i in range(10):
        time.sleep(1)
        print(i+1,"done")
        return_value, image = camera.read()
        if not os.path.exists('known'):
            os.makedirs('known')
        cv2.imwrite('known\opencv'+str(i)+'.png', image)
    del(camera)

    print("Encoding...")
    for filename in os.listdir(knownImgPath):
        known=face_recognition.load_image_file(os.path.join(knownImgPath, filename))
        try:
            knownEncoding=face_recognition.face_encodings(known)[0]
            knownFace.append(knownEncoding)
        except IndexError:
            print("Failed to encode. Trying again.")
            capture_image()

    print("Encoded")

def find_images(lower,upper,files,knownFace,pid):
    counter1=0
    counter2=0
    for i in range(lower,upper):
        filename=files[i]
        faceCascade=cv2.CascadeClassifier("haarcascade\haarcascade_frontalface_alt.xml")
        img=cv2.imread(filename)
        gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
        faces=faceCascade.detectMultiScale(gray,1.05,6,minSize=(30,30))
        print("Found {} face(s) in {}".format(len(faces),filename))

        for (x, y, w, h) in faces:
            cv2.rectangle(img, (x, y), (x+w, y+h), (0, 0, 255), 2)
            faces = img[y:y + h, x:x + w]
            if pid==1: 
                if not os.path.exists('faces\\temp1'):
                    os.makedirs('faces\\temp1')

                cv2.imwrite('faces\\temp1\\temp_face'+str(counter1)+'.jpg', faces)

            elif pid==2: 
                if not os.path.exists('faces\\temp2'):
                    os.makedirs('faces\\temp2')

                cv2.imwrite('faces\\temp2\\temp_face'+str(counter2)+'.jpg', faces)

            if pid==1: unknown=face_recognition.load_image_file('faces\\temp1\\temp_face'+str(counter1)+'.jpg')
            elif pid==2: unknown=face_recognition.load_image_file('faces\\temp2\\temp_face'+str(counter2)+'.jpg')

            unknownEncodings=face_recognition.face_encodings(unknown)

            if len(unknownEncodings)>0:
                result=face_recognition.compare_faces(knownFace,unknownEncodings[0],0.55)
            
                if result.count(True)>=5:
                    if not os.path.exists('my_photos'):
                        os.makedirs('my_photos')
                    copy(filename,'my_photos')
                    break

                if pid==1: counter1+=1         
                elif pid==2: counter2+=1                        

if __name__ =='__main__':
    start= time.time()
    knownFace=[]
    files=[]
    directory=input("Enter directory: ")
    for filename in os.listdir(directory):
        f = os.path.join(directory, filename)
        if (f.lower().endswith(".jpg") or f.lower().endswith(".png") or f.lower().endswith(".jpeg")):
            files.append(f)

    print(files)
    capture_image()
    p1=multiprocessing.Process(target=find_images, args=(0,int(len(files)/2),files,knownFace,1))
    p2=multiprocessing.Process(target=find_images, args=(int(len(files)/2),len(files),files,knownFace,2))
    p1.start()
    p2.start()
    p1.join()
    p2.join()
    shutil.rmtree('faces', ignore_errors=True)
    shutil.rmtree('known', ignore_errors=True)
    end=time.time()
    print(end-start)