1.
Detection
import os,cv2,numpy as np
if not os.path.exists('dataset'):
os.makedirs('dataset')
face = os.listdir('dataset')
facedetect=cv2.CascadeClassifier(cv2.data.haarcascades +
'haarcascade_frontalface_default.xml')
recognizer = cv2.face.LBPHFaceRecognizer_create()
def dataset():
nama = simpledialog.askstring("Input Nama", "Masukkan Nama Anda!")
if not nama:
messagebox.showinfo("Informasi", "Isikan Nama Anda!")
return
cam = cv2.VideoCapture(1)
cam.set(3, 360) # set lebar video
cam.set(4, 270) # set tinggi video
if not os.path.exists('dataset/'+nama):
os.makedirs('dataset/'+nama.lower())
count = 0
while(True):
ret, img = cam.read()
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
faces = facedetect.detectMultiScale(gray, 1.3, 5)
for (x,y,w,h) in faces:
cv2.rectangle(img, (x, y), (x + w, y + h), (0, 0, 255), 2)
count += 1
cv2.imwrite("dataset/" + str(nama) + '/' + str(count) + ".jpg", gray[y:y+h,x:x+w])
cv2.putText(img, str(count), (x + 5, y - 5), cv2.FONT_HERSHEY_SIMPLEX, 1,
(255, 255, 255), 2)
cv2.imshow('img',img)
k = cv2.waitKey(100) & 0xff # Tekan 'ESC' untuk keluar dari Video
if k == 27:
break
if count == 30: # Mengambil foto wajah sebanyak 30 dan menghentikan Video
break
cam.release()
cv2.destroyAllWindows()
messagebox.showinfo("Informasi", "Dataset "+nama+" telah berhasil
ditambahkan!")
def training():
face = os.listdir('dataset')
facemodels=[]
ids = []
cnt=0
for v in face:
for f in os.listdir(os.path.join('dataset/', v)):
PIL_img = Image.open(os.path.join('dataset/', v, f)).convert('L') # Mengubah ke
Grayscale
img_numpy = np.array(PIL_img, 'uint8')
faces = facedetect.detectMultiScale(img_numpy)
for (x, y, w, h) in faces:
facemodels.append(img_numpy[y:y + h, x:x + w])
ids.append(cnt)
cnt+=1
recognizer.train(facemodels,np.array(ids))
recognizer.write('models.yml')
messagebox.showinfo("Informasi", "Training telah berhasil!")
root = Tk()
root.title("Pengambilan Data")
root.geometry('480x360')
root.eval('tk::PlaceWindow . center')
picPanel = Label(root, text='Preprocessed',borderwidth=2, relief='groove')
picPanel.place(x=100, y=25, width=360, height=270)
btn_dataset = Button(root, text="Dataset", command=dataset)
btn_dataset.place(x=15, y=25, width=75, height=20)
btn_training = Button(root, text="Training", command=training)
btn_training.place(x=15, y=50, width=75, height=20)
keterangan = Label(root,text="Prediksi")
keterangan.place(x=100, y=300, width=50, height=20)
person = Label(root,borderwidth=2, relief='groove')
person.place(x=150, y=300, width=100, height=20)
root.mainloop()
2. Training Data
from tkinter import *
from tkinter import simpledialog,messagebox
from PIL import ImageTk, Image
import os,cv2,numpy as np
if not os.path.exists('dataset'):
os.makedirs('dataset')
face = os.listdir('dataset')
facedetect=cv2.CascadeClassifier(cv2.data.haarcascades +
'haarcascade_frontalface_default.xml')
recognizer = cv2.face.LBPHFaceRecognizer_create()
def dataset():
nama = simpledialog.askstring("Input Nama", "Masukkan Nama Anda!")
if not nama:
messagebox.showinfo("Informasi", "Isikan Nama Anda!")
return
cam = cv2.VideoCapture(1)
cam.set(3, 360) # set lebar video
cam.set(4, 270) # set tinggi video
if not os.path.exists('dataset/'+nama):
os.makedirs('dataset/'+nama.lower())
count = 0
while(True):
ret, img = cam.read()
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
faces = facedetect.detectMultiScale(gray, 1.3, 5)
for (x,y,w,h) in faces:
cv2.rectangle(img, (x, y), (x + w, y + h), (0, 0, 255), 2)
count += 1
cv2.imwrite("dataset/" + str(nama) + '/' + str(count) + ".jpg", gray[y:y+h,x:x+w])
cv2.putText(img, str(count), (x + 5, y - 5), cv2.FONT_HERSHEY_SIMPLEX, 1,
(255, 255, 255), 2)
cv2.imshow('img',img)
k = cv2.waitKey(100) & 0xff # Tekan 'ESC' untuk keluar dari Video
if k == 27:
break
if count == 30: # Mengambil foto wajah sebanyak 30 dan menghentikan Video
break
cam.release()
cv2.destroyAllWindows()
messagebox.showinfo("Informasi", "Dataset "+nama+" telah berhasil
ditambahkan!")
def training():
face = os.listdir('dataset')
facemodels=[]
ids = []
cnt=0
for v in face:
for f in os.listdir(os.path.join('dataset/', v)):
PIL_img = Image.open(os.path.join('dataset/', v, f)).convert('L') # Mengubah ke
Grayscale
img_numpy = np.array(PIL_img, 'uint8')
faces = facedetect.detectMultiScale(img_numpy)
for (x, y, w, h) in faces:
facemodels.append(img_numpy[y:y + h, x:x + w])
ids.append(cnt)
cnt+=1
recognizer.train(facemodels,np.array(ids))
recognizer.write('models.yml')
messagebox.showinfo("Informasi", "Training telah berhasil!")
root = Tk()
root.title("Pengambilan Data")
root.geometry('480x360')
root.eval('tk::PlaceWindow . center')
picPanel = Label(root, text='Preprocessed',borderwidth=2, relief='groove')
picPanel.place(x=100, y=25, width=360, height=270)
btn_dataset = Button(root, text="Dataset", command=dataset)
btn_dataset.place(x=15, y=25, width=75, height=20)
btn_training = Button(root, text="Training", command=training)
btn_training.place(x=15, y=50, width=75, height=20)
keterangan = Label(root,text="Prediksi")
keterangan.place(x=100, y=300, width=50, height=20)
person = Label(root,borderwidth=2, relief='groove')
person.place(x=150, y=300, width=100, height=20)
root.mainloop()
