Search code examples
tensorflowmachine-learningkerasimage-augmentation

The colour of my images get changed after augmentation

I am using Keras and ImageDataGenerator to augment the images in a directory and save it in another folder. After Augmentation the colour of the images turn bluish negativish and i dont want the colour to change.

Before

After

train_datagen = ImageDataGenerator(
                                   rotation_range = 10,
                                   shear_range = 0.15,
                                   zoom_range = 0.15,
                                   horizontal_flip = True,
                                    brightness_range=[0.5,1.5])
train_batch_size = 32
val_batch_size = 32
import warnings
warnings.filterwarnings('ignore')

for img in imgs:
    img=cv2.imread(train_path+'real'+"\\"+img)
    x = img_to_array(img)
    print(x.shape)
    x = x.reshape((1,) + x.shape)
    print(x)
    break

    i = 0
    for batch in train_datagen.flow(x, batch_size=1, save_to_dir =r'C:\Users\Shashwat Goyal\Desktop\augment', save_prefix ='people2', save_format='jpg'):
        i+=1
        if i>4:
            break
Solution

  • You are using cv2 to read in your images. cv2 formats images as BGR versus RGB. To solve your problem use code

    img=cv2.imread(train_path+'real'+"\\"+img)
x= cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

    Note you do not need the code

     x = img_to_array(img)

    because a cv2 image is already an array