def setUp(self):
np.random.seed(1234)
self.data_input_3D = np.random.random((2, 64, 56, 48))
self.data_input_2D = np.random.random((2, 64, 56))
self.d_3D = augment_gamma(np.copy(self.data_input_2D),
gamma_range=(0.2, 1.2),
per_channel=False)
def __call__(self, **data_dict):
data_dict[self.data_key] = augment_gamma(
data_dict[self.data_key],
self.gamma_range,
self.invert_image,
per_channel=self.per_channel,
retain_stats=self.retain_stats)
return data_dict
def __call__(self, **data_dict):
for b in range(len(data_dict[self.data_key])):
if np.random.uniform() < self.p_per_sample:
data_dict[self.data_key][b] = augment_gamma(data_dict[self.data_key][b], self.gamma_range,
self.invert_image,
per_channel=self.per_channel,
retain_stats=self.retain_stats)
return data_dict
def gamma_augmentation_generator(generator, gamma_range=(0.5, 2), invert_image=False):
warn("using deprecated generator brightness_augmentation_by_multiplication_generator", Warning)
# augments by shifting the gamma value as in gamma correction (https://en.wikipedia.org/wiki/Gamma_correction)
for data_dict in generator:
data_dict['data'] = augment_gamma(data_dict['data'], gamma_range, invert_image)
yield data_dict
def __call__(self, **data_dict):
data_dict['data'] = augment_gamma(data_dict['data'], self.gamma_range,
self.invert_image)
return data_dict
def __call__(self, **data_dict):
data_dict[self.data_key] = augment_gamma(data_dict[self.data_key], self.gamma_range, self.invert_image,
per_channel=self.per_channel, retain_stats=self.retain_stats)
return data_dict
