import images from folder python to numpy array list

Question

I have a folder that contains 10000 images, and 3 subfolders , each folder contains different number of images. I want to import a small portion of these images for training, that the limited size i chose manually each time i want to pick a portion of the data. I have already this python code :

train_dir = 'folder/train/' # This folder contains 10.000 images and 3 subfolders , each folder contains different number of images

from tqdm import tqdm
def get_data(folder):
    """
    Load the data and labels from the given folder.
    """
    X = []
    y = []
    for folderName in os.listdir(folder):
        if not folderName.startswith('.'):
            if folderName in   ['Name1']:
                label = 0
            elif folderName in ['Name2']:
                label = 1
            elif folderName in ['Name3']:
                label = 2
            else:
                label = 4
            for image_filename in tqdm(os.listdir(folder + folderName)):
                img_file = cv2.imread(folder + folderName + '/' + image_filename)
                if img_file is not None:
                    img_file = skimage.transform.resize(img_file, (imageSize, imageSize, 1))
                    img_arr = np.asarray(img_file)
                    X.append(img_arr)
                    y.append(label)
    X = np.asarray(X) # Keras only accepts data as numpy arrays 
    y = np.asarray(y)
    return X,y


X_test, y_test= get_data(train_dir)
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X_test, y_test, test_size=0.2)

i want to specify Size parameter so that i can choose the number of images to import. the number of imported images from each subfolder should be equal

It seems like what you need is Keras ImageDataGenerator class with flow_from_directory. keras.io/preprocessing/image/#imagedatagenerator-class — kuzand
– kuzand, Commented Feb 28, 2019 at 13:15
is it possible to specify the number of images imported from a folder using ImageDataGenerator ? if so , how ? — Mohamed Berrimi
– Mohamed Berrimi, Commented Mar 3, 2019 at 14:08

Bsquare ℬℬ · Accepted Answer · 2019-02-28 10:21:55Z

1

You can read and store every paths from each folder in a separate list and select equal number of them.

folder1_files = []
for root, dirs, files in os.walk('path/folder1', topdown=False):
    for i in files:
        folder1_files.append("path/folder1/"+i)

to select:

train = folder1[:n] + folder2[:n] + folder3[:n]

n - number of images from each folder

edited Feb 28, 2019 at 10:21

Bsquare ℬℬ

4,49711 gold badges27 silver badges44 bronze badges

answered Feb 28, 2019 at 10:16

newlearnershiv

3601 silver badge9 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Mohamed Berrimi Over a year ago

Thanks, but i want to upload with the code structure i mentioned above ( there encodings and many things .. ), is it possible to do that ?

Collectives™ on Stack Overflow

import images from folder python to numpy array list

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related