Face Detection Project code using Python

Introduction

In our system the face detection is performed on frames acquired from real time video.
Then a face classification method that uses Convolutional Neural Network is integrated in the system.
The system use standard images, and detection algorithm to train itself, extract faces, and detect the face.

Objectives of Face Detection

To implement a face net system to produce facial Detection application.

To design a system that can detect and recognize faces in real time.

Functional requirements of Face Detection

Can detect multiple faces in a real-time video.
Acquiring frames from real time video.
Train the system module.
Create classifier files of the data set.
User friendly interface.
Enter new data or check user.
Robust and machine friendly.

Enter user Id/Name.
Collect Data Set.
Train Data model.
Create classifiers.
Select user.
Recognize face.
Show Id/Name.

Non-Functional Requirements of Face Detection

Robustness
Stability
Accuracy
Security

Tools and Technologies used in Face Detection

Operating System : Linux (Ubuntu) 64 bit / Windows 64 bit

Hardware : 4 GB-RAM, Webcam

Programming Language : Python

Computer Vision Library : OpenCV

GUI Library: Tkinter

Algorithm : LBPH (Local Binary Pattern Histogram)

Modules of Face Detection

Our system have following Modules.

App_GUI: Runs the application of user interface.
Collect data set: Collect frames from runtime video.
Trainer: Train the dataset and create classifiers.
Face Recognizer: By using data classifiers recognizes the person in real-time and displays the Id.

Code

create_classifier.py

import cv2
import numpy as np
from PIL import Image
import os




def train_classifer(name):
    path = os.path.join(os.getcwd()+"/data/"+name+"/")

    faces = []
    ids = []
    labels = []
    pictures = {}


   
    for root,dirs,files in os.walk(path):
            pictures = files


    for pic in pictures :

            imgpath = path+pic
            img = Image.open(imgpath).convert('L')
            imageNp = np.array(img, 'uint8')
            id = int(pic.split(name)[0])
            faces.append(imageNp)
            ids.append(id)

    ids = np.array(ids)

    clf = cv2.face.LBPHFaceRecognizer_create()
    clf.train(faces, ids)
    clf.write("./data/classifiers/"+name+"_classifier.xml")

import cv2

import numpy as np

from PIL import Image

import os

def train_classifer(name):

path = os.path.join(os.getcwd()+"/data/"+name+"/")

faces = []

ids = []

labels = []

pictures = {}

for root,dirs,files in os.walk(path):

pictures = files

for pic in pictures :

imgpath = path+pic

img = Image.open(imgpath).convert('L')

imageNp = np.array(img, 'uint8')

id = int(pic.split(name)[0])

faces.append(imageNp)

ids.append(id)

ids = np.array(ids)

clf = cv2.face.LBPHFaceRecognizer_create()

clf.train(faces, ids)

clf.write("./data/classifiers/"+name+"_classifier.xml")

create_dataset.py

import cv2
import os

def start_capture(name):
        path = "./data/" + name
        num_of_images = 0
        detector = cv2.CascadeClassifier("./data/haarcascade_frontalface_default.xml")
        try:
            os.makedirs(path)
        except:
            print('Directory Already Created')
        vid = cv2.VideoCapture(0)
        while True:

            ret, img = vid.read()
            new_img = None
            grayimg = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
            face = detector.detectMultiScale(image=grayimg, scaleFactor=1.1, minNeighbors=5)
            for x, y, w, h in face:
                cv2.rectangle(img, (x, y), (x+w, y+h), (0, 0, 0), 2)
                cv2.putText(img, "Face Detected", (x, y-5), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 0, 255))
                cv2.putText(img, str(str(num_of_images)+" images captured"), (x, y+h+20), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 0, 255))
                new_img = img[y:y+h, x:x+w]
            cv2.imshow("FaceDetection", img)
            key = cv2.waitKey(1) & 0xFF


            try :
                cv2.imwrite(str(path+"/"+str(num_of_images)+name+".jpg"), new_img)
                num_of_images += 1
            except :

                pass
            if key == ord("q") or key == 27 or num_of_images > 310:
                break
        cv2.destroyAllWindows()
        return num_of_images

import cv2

import os

def start_capture(name):

path = "./data/" + name

num_of_images = 0

detector = cv2.CascadeClassifier("./data/haarcascade_frontalface_default.xml")

try:

os.makedirs(path)

except:

print('Directory Already Created')

vid = cv2.VideoCapture(0)

while True:

ret, img = vid.read()

new_img = None

grayimg = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

face = detector.detectMultiScale(image=grayimg, scaleFactor=1.1, minNeighbors=5)

for x, y, w, h in face:

cv2.rectangle(img, (x, y), (x+w, y+h), (0, 0, 0), 2)

cv2.putText(img, "Face Detected", (x, y-5), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 0, 255))

cv2.putText(img, str(str(num_of_images)+" images captured"), (x, y+h+20), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 0, 255))

new_img = img[y:y+h, x:x+w]

cv2.imshow("FaceDetection", img)

key = cv2.waitKey(1) & 0xFF

try :

cv2.imwrite(str(path+"/"+str(num_of_images)+name+".jpg"), new_img)

num_of_images += 1

except :

pass

if key == ord("q") or key == 27 or num_of_images > 310:

break

cv2.destroyAllWindows()

return num_of_images

Detector.py

import cv2
from time import sleep
from PIL import Image 

def main_app(name):
        
        face_cascade = cv2.CascadeClassifier('./data/haarcascade_frontalface_default.xml')
        recognizer = cv2.face.LBPHFaceRecognizer_create()
        recognizer.read(f"./data/classifiers/{name}_classifier.xml")
        cap = cv2.VideoCapture(0)
        pred = 0
        while True:
            ret, frame = cap.read()
            gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
            faces = face_cascade.detectMultiScale(gray,1.3,5)

            for (x,y,w,h) in faces:


                roi_gray = gray[y:y+h,x:x+w]

                id,confidence = recognizer.predict(roi_gray)
                confidence = 100 - int(confidence)
                pred = 0
                if confidence > 50:
                          pred += +1
                            text = name.upper()
                            font = cv2.FONT_HERSHEY_PLAIN
                            frame = cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
                            frame = cv2.putText(frame, text, (x, y-4), font, 1, (0, 255, 0), 1, cv2.LINE_AA)

                else:   
                            pred += -1
                            text = "UnknownFace"
                            font = cv2.FONT_HERSHEY_PLAIN
                            frame = cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 0, 255), 2)
                            frame = cv2.putText(frame, text, (x, y-4), font, 1, (0, 0,255), 1, cv2.LINE_AA)

            cv2.imshow("image", frame)


            if cv2.waitKey(20) & 0xFF == ord('q'):
                print(pred)
                if pred > 0 : 
                    dim =(124,124)
                    img = cv2.imread(f".\\data\\{name}\\{pred}{name}.jpg", cv2.IMREAD_UNCHANGED)
                    resized = cv2.resize(img, dim, interpolation = cv2.INTER_AREA)
                    cv2.imwrite(f".\\data\\{name}\\50{name}.jpg", resized)
                    Image1 = Image.open(f".\\2.png") 
                      
                   Image1copy = Image1.copy() 
                    Image2 = Image.open(f".\\data\\{name}\\50{name}.jpg") 
                    Image2copy = Image2.copy() 
                      
                    Image1copy.paste(Image2copy, (195, 114)) 
                      
                    Image1copy.save("end.png") 
                    frame = cv2.imread("end.png", 1)

                    cv2.imshow("Result",frame)
                    cv2.waitKey(5000)
                break


        cap.release()
        cv2.destroyAllWindows()

import cv2

from time import sleep

from PIL import Image

def main_app(name):

face_cascade = cv2.CascadeClassifier('./data/haarcascade_frontalface_default.xml')

recognizer = cv2.face.LBPHFaceRecognizer_create()

recognizer.read(f"./data/classifiers/{name}_classifier.xml")

cap = cv2.VideoCapture(0)

pred = 0

while True:

ret, frame = cap.read()

gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

faces = face_cascade.detectMultiScale(gray,1.3,5)

for (x,y,w,h) in faces:

roi_gray = gray[y:y+h,x:x+w]

id,confidence = recognizer.predict(roi_gray)

confidence = 100 - int(confidence)

pred = 0

if confidence > 50:

pred += +1

text = name.upper()

font = cv2.FONT_HERSHEY_PLAIN

frame = cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)

frame = cv2.putText(frame, text, (x, y-4), font, 1, (0, 255, 0), 1, cv2.LINE_AA)

else:

pred += -1

text = "UnknownFace"

font = cv2.FONT_HERSHEY_PLAIN

frame = cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 0, 255), 2)

frame = cv2.putText(frame, text, (x, y-4), font, 1, (0, 0,255), 1, cv2.LINE_AA)

cv2.imshow("image", frame)

if cv2.waitKey(20) & 0xFF == ord('q'):

print(pred)

if pred > 0 :

dim =(124,124)

img = cv2.imread(f".\\data\\{name}\\{pred}{name}.jpg", cv2.IMREAD_UNCHANGED)

resized = cv2.resize(img, dim, interpolation = cv2.INTER_AREA)

cv2.imwrite(f".\\data\\{name}\\50{name}.jpg", resized)

Image1 = Image.open(f".\\2.png")

Image1copy = Image1.copy()

Image2 = Image.open(f".\\data\\{name}\\50{name}.jpg")

Image2copy = Image2.copy()

Image1copy.paste(Image2copy, (195, 114))

Image1copy.save("end.png")

frame = cv2.imread("end.png", 1)

cv2.imshow("Result",frame)

cv2.waitKey(5000)

break

cap.release()

cv2.destroyAllWindows()

Face Detection Project code using Python

Code

Download full project code of Face Detection in Python

Download Face Detection PPT Presentation of Project

Related Posts:

Leave a Comment Cancel Reply