dataaug

import cv2
import numpy as np
import os
from tqdm import tqdm
from sklearn.model_selection import train_test_split
import scipy.misc
from keras.layers import Dense, Flatten, Input, Conv2D, Conv2DTranspose, MaxPooling2D, UpSampling2D
from keras.models import Model
from PIL import Image
from keras.utils import to_categorical
from sklearn.utils import shuffle
from keras.preprocessing.image import ImageDataGenerator
from keras.preprocessing.image import array_to_img, img_to_array, load_img
gen = ImageDataGenerator(
    rotation_range=20,
    width_shift_range=0.2,
    height_shift_range=0.2,
    horizontal_flip=True
)

APPLE_DIRECTORY = "./apple/"

apple_images = []


for apple_dir in tqdm(os.listdir(APPLE_DIRECTORY)):
    for apple in os.listdir(APPLE_DIRECTORY + '/' + apple_dir):
         if apple.endswith('_crop.png'):
            img = cv2.imread(APPLE_DIRECTORY + '/' + apple_dir + '/' + apple)
            img = cv2.resize(img, (92,92), interpolation = cv2.INTER_AREA)
            #img = img_to_array(load_img(APPLE_DIRECTORY + '/' + apple_dir + '/' + apple))
            img = img_to_array(img)
            img = img.reshape((1,) + img.shape)

            apple_images.append(img)
            #image= apple_images.reshape((1,)+ apple_images.shape)
output_path = './apple_aug/apple_random{}.png'
count =10
# let's create infinite flow of images
images_flow = gen.flow(img, batch_size=100)
for i, new_images in enumerate(images_flow):
    # we access only first image because of batch_size=1
    new_image = array_to_img(new_images[0], scale=True)
    new_image.save(output_path.format(i + 1))
    if i >= count:
        break