I want to apply a OpenCV colormap filter in all my Train images, but i'm using the flow_from_directory method as follows:
train_datagen = ImageDataGenerator()
train_generator = train_datagen.flow_from_directory(TRAIN_DIR,
target_size=(224, 224),
batch_size=50,
class_mode='binary')
How can I add the colormap in this case?
You can pass a preprocessing function to ImageDataGenerator.
def cmap_transf(image):
return cv2.applyColorMap(image.astype(np.uint8), cv2.COLORMAP_JET)
train_generator = ImageDataGenerator(preprocessing_function=cmap_transf)
Full working example:
import cv2
import numpy as np
from tensorflow.keras.preprocessing.image import ImageDataGenerator
import os
import matplotlib.pyplot as plt
os.makedirs('test', exist_ok=True)
cv2.imwrite('test/img.jpg', np.random.randint(0, 256, (200, 200, 3)))
def cmap_transf(image):
return cv2.applyColorMap(image.astype(np.uint8), cv2.COLORMAP_JET)
train_generator = ImageDataGenerator(preprocessing_function=cmap_transf)
plt.imshow(next(iter(train_generator.
flow_from_directory('.',
classes=['test'])))[0][0, ...])
os.unlink('test/img.jpg')
os.rmdir('test')
Original:
Preprocessed:
