import cv2
import os
import numpy as np

import matplotlib.pyplot as plt

from sklearn.model_selection import train_test_split
from sklearn.datasets import fetch_lfw_people
from sklearn.metrics import classification_report
from sklearn.decomposition import PCA
from sklearn.neural_network import MLPClassifier

subjects = ["0", "1", "2"]

# Prepare training data
data_folder_path = 'C:/Users/MUICT/Desktop/jupyter/face/'
dirs = os.listdir(data_folder_path)

faces = []
labels = []

face_cascade = cv2.CascadeClassifier('D:/opencv/sources/data/hogcascades/haarcascade_frontalface_default.xml')

for dir_name in dirs:
    subject_dir_path = data_folder_path + dir_name
    print(subject_dir_path)
    subject_images_names = os.listdir(subject_dir_path)

    for image_name in subject_images_names:
        image_path = subject_dir_path + "/" + image_name
        image = cv2.imread(image_path)
        gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
        facesL = face_cascade.detectMultiScale(gray, scaleFactor=1.2, minNeighbors=5);
        if (len(facesL) > 0):
            (x, y, w, h) = facesL[0]
            faces.append(cv2.resize(gray[y:y+w, x:x+h],(70,70)))
            labels.append(int(dir_name))

print("Total faces: ", len(faces))
print("Total labels: ", len(labels))

#recognizer = cv2.face.EigenFaceRecognizer_create()
recognizer = cv2.face.LBPHFaceRecognizer_create()

#print(labels)
recognizer.train(faces, np.array(labels))

# Predict
test = cv2.imread('C:/Users/MUICT/Desktop/jupyter/test.jpg')
testG = cv2.cvtColor(test, cv2.COLOR_BGR2GRAY)
testF = face_cascade.detectMultiScale(testG, scaleFactor=1.2, minNeighbors=5);
(x, y, w, h) = testF[0]
testFF = cv2.resize(testG[y:y+w, x:x+h],(70,70))
label, confidence = recognizer.predict(testFF)
print(label)