def main(datadir, convert_dir, crop_size):
try:
os.mkdir(convert_dir)
except OSError:
pass
filenames = data_util.get_image_files(datadir)
print('Resizing images in {} to {}'.format(datadir, convert_dir))
n = len(filenames)
batch_size = 500
batches = n // batch_size + 1
p = Pool()
args = []
for f in filenames:
args.append((convert_size, (datadir, convert_dir, f, crop_size)))
for i in range(batches):
print('batch {:>2} / {}'.format(i + 1, batches))
p.map(convert, args[i * batch_size : (i + 1) * batch_size])
p.close()
p.join()
print('Done')
def main(datadir, convert_dir, crop_size):
try:
os.mkdir(convert_dir)
except OSError:
pass
filenames = data_util.get_image_files(datadir)
print('Resizing images in {} to {}'.format(datadir, convert_dir))
n = len(filenames)
batch_size = 500
batches = n // batch_size + 1
p = Pool()
args = []
for f in filenames:
args.append((convert_size, (datadir, convert_dir, f, crop_size)))
for i in range(batches):
print('batch {:>2} / {}'.format(i + 1, batches))
p.map(convert, args[i * batch_size:(i + 1) * batch_size])
p.close()
p.join()
print('Done')
import numpy as np
from time import time
import pdb
import skimage
import matplotlib.pyplot as plt
import data_util
DATA_DIR = "converted"
files = data_util.get_image_files(DATA_DIR)
images = data_util.load_images(files)
MEAN = data_util.compute_mean(files)
STD = data_util.compute_std(files)
images_normalized = []
for img in images:
img = img - MEAN[:, np.newaxis, np.newaxis]
img = img / STD[:, np.newaxis, np.newaxis]
images_normalized.append(img)
images_normalized = np.array(images_normalized)
original_augmented = data_util.parallel_augment(images)
normalized_augmented = data_util.parallel_augment(images_normalized)
original = images[3]
normalized = images_normalized[3]
original = original.transpose(1, 2, 0)
normalized = normalized.transpose(1, 2, 0)
import numpy as np
import cPickle
import data_util
DATA_DIR = '/nikel/dhpark/fundus/kaggle/original/training/train_medium'
files = data_util.get_image_files(DATA_DIR)
mean = data_util.compute_mean_across_channels(files)
std = data_util.compute_std_across_channels(files)
print("computing done")
print("dumping...")
mean.dump("mean.dat")
std.dump("std.dat")
import numpy as np
from time import time
import data
import data_util
from matplotlib import pyplot as plt
aug_params = {
'zoom_range': (1 / 1.15, 1.15),
'rotation_range': (0, 360),
'shear_range': (0, 0),
'translation_range': (-40, 40),
'do_flip': True,
'allow_stretch': True,
}
files = data_util.get_image_files('testing')
X = data_util.load_images(files)
mean, std = data_util.compute_mean_and_std(files)
print(mean, std)
print("Number of images: {}".format(len(X)))
# start = time()
# result = data.batch_perturb_and_augment(X, 500, 500, aug_params=aug_params, sigma=0.5)
# end = time()
# print("Processing without parallelization took {} seconds".format(end - start))
#start = time()
#result = data.parallel_perturb_and_augment(X, 500, 500, aug_params=aug_params, sigma=0.5)
#result = data.parallel_perturb_and_augment(X, 500, 500)
#end = time()
import data_util
from matplotlib import pyplot as plt
aug_params = {
'zoom_range': (1 / 1.15, 1.15),
'rotation_range': (0, 360),
'shear_range': (0, 0),
'translation_range': (-40, 40),
'do_flip': True,
'allow_stretch': True,
}
files = data_util.get_image_files('testing')
X = data_util.load_images(files)
mean, std = data_util.compute_mean_and_std(files)
print(mean, std)
print("Number of images: {}".format(len(X)))
# start = time()
# result = data.batch_perturb_and_augment(X, 500, 500, aug_params=aug_params, sigma=0.5)
# end = time()
# print("Processing without parallelization took {} seconds".format(end - start))
#start = time()
#result = data.parallel_perturb_and_augment(X, 500, 500, aug_params=aug_params, sigma=0.5)
#result = data.parallel_perturb_and_augment(X, 500, 500)
#end = time()