ImageDataGenerator loaded images lost lots of features

TensorFlow General Discussion
, ,
1

Hello there,

I am using Keras ImageDataGenerator in the project. After not achieving desired validation losses, I found something weird.

In here I have original image of this;
(Originally 1200x1200)

doğru
doğru1279×1279 22 KB

I need to feed 100x100 input layered model with this image. Therefore, I loaded the image with;

 train_datagen = ImageDataGenerator(
        validation_split=0,
        rotation_range=5,  # rotation
        horizontal_flip=True,  # horizontal flip
    )
train_generator = train_datagen.flow_from_directory(
        directory=generated_images_folder + "training\\" + user.username + "\\train",
        color_mode="rgb",
        batch_size=batch_size,
        # target_size=(100, 100),
        class_mode="categorical",
        shuffle=False
    )

With or without target_size=(100,100) argument.
This is 100x100;

100_100
100_100640×480 6.09 KB

And this is 255x255;

255_255
255_255640×480 11.5 KB

After loading this image, I rescaled this with cv2;

image = cv2.resize(x[0], (100, 100))

Here is the result;

255_to_100
255_to_100640×480 6.21 KB

So, question is that why ImageDataGenerator loaded image is loosing that much of the image? Is there a bug here?

3

You can change the interpolation arg as by default it is nearest.

4

Good point. However, none of tried interpolation method gave me clean result. They either blur the image or make several pixels white, since most of the images are white.

5

Internally are just PIL/Pillow calls so you can test the behaviour with the library directly:

https://github.com/keras-team/keras/blob/master/keras/preprocessing/image.py#L381

6

Thank you again. I realized that too. PIL doesn’t allow us to use none of the interpolation methods despite of cv2, I guess. This may be the why cv2 gives cleaner results.

7

I cannot replicate your problem on Colab with PIL:

from PIL import Image
import tensorflow as tf

im = Image.open("/content/doğru.jpg")
b = im.resize((100,100))
display(b)
9

Hi, again.
Here is the reproducing example;

from PIL import Image
import tensorflow as tf
import numpy as np
from matplotlib import pyplot as plt

im = Image.open("/content/deneme/class1/doğru.jpg")
b = im.resize((100,100))
plt.imshow(b)
plt.show()

train_datagen = tf.keras.preprocessing.image.ImageDataGenerator(
        rescale=1 / 255.0,
    )
train_generator = train_datagen.flow_from_directory(
        directory="/content/deneme",
        color_mode="rgb",
        batch_size=32,
        target_size=(100, 100),
        class_mode="categorical",
        shuffle=False
    )

x, y = train_generator.__next__()
plt.imshow(x[0])
plt.show()

In here, I would like to point out that resize with PIL is okay but image coming from generator is ruined.

11

You need to use the same filter type: e.g. interpolation='bicubic'

See more

13

Thank you, you’re correct. I wonder why default interpolation is nearest in tf, while it is bicubic in PIL (see below)

    def resize(self, size, resample=BICUBIC, box=None, reducing_gap=None):
        """
        Returns a resized copy of this image.
         ...
        :param resample: An optional resampling filter.  This can be
           one of :py:data:`PIL.Image.NEAREST`, :py:data:`PIL.Image.BOX`,
           :py:data:`PIL.Image.BILINEAR`, :py:data:`PIL.Image.HAMMING`,
           :py:data:`PIL.Image.BICUBIC` or :py:data:`PIL.Image.LANCZOS`.
           Default filter is :py:data:`PIL.Image.BICUBIC`.
           If the image has mode "1" or "P", it is
           always set to :py:data:`PIL.Image.NEAREST`.
           See: :ref:`concept-filters`.

Is it because of better performance? (Shown in; Concepts - Pillow (PIL Fork) 10.4.0 documentation)

14

It is how it Is used in that Keras API as in other TF API the default is different.

See tf.image.resize  |  TensorFlow v2.16.1

15

Fair enough! Thanks again for your help.

16

Honestly it could be nice to have a consistent default behavior. Probably you could open a ticket in the Keras repo.

1 Like