global args
args = parse_args()
interp_list = [cv.INTER_NEAREST, cv.INTER_LINEAR, cv.INTER_CUBIC, cv.INTER_LANCZOS4]
def return_raw_image(dataset):
dataset_raw = []
for image_features in dataset:
image_raw = image_features['image'].numpy()
image = tf.image.decode_jpeg(image_raw)
dataset_raw.append(image)
return dataset_raw
fg_dataset = tfrecord_creator.read("fg", "./data/tfrecord/")
bg_dataset = tfrecord_creator.read("bg", "./data/tfrecord/")
a_dataset = tfrecord_creator.read("a", "./data/tfrecord/")
fg_dataset = list(fg_dataset)
bg_dataset = list(bg_dataset)
a_dataset = list(a_dataset)
# fg_raw = return_raw_image(fg_dataset)
# bg_raw = return_raw_image(bg_dataset)
# a_raw = return_raw_image(a_dataset)
def get_raw(type_of_dataset, count):
if type_of_dataset == 'fg':
temp = fg_dataset[count]['image']
channels=3
elif type_of_dataset == 'bg':
temp = bg_dataset[count]['image']
}
def return_raw_image(dataset):
dataset_raw = []
for image_features in dataset:
image_raw = image_features['image'].numpy()
image = tf.image.decode_jpeg(image_raw)
dataset_raw.append(image)
return dataset_raw
# bg_dataset = tfrecord_creator.read("bg", "./data/tfrecord/")
# bg_dataset = tfrecord_creator.read("bg", "../data/bg/")
bg_dataset = tfrecord_creator.read("bg", "/content/bg/")
bg_dataset = list(bg_dataset)
print("___________________")
print(len(bg_dataset))
print("___________________")
def get_raw(type_of_dataset, count):
if type_of_dataset == 'fg':
temp = fg_dataset[count]['image']
channels = 3
elif type_of_dataset == 'bg':
temp = bg_dataset[count]['image']
channels = 3
else:
temp = a_dataset[count]['image']