Permalink
Branch: master
Find file Copy path
2331 lines (1824 sloc) 100 KB
"""
Augmenters that perform simple arithmetic changes.
Do not import directly from this file, as the categorization is not final.
Use instead::
from imgaug import augmenters as iaa
and then e.g.::
seq = iaa.Sequential([iaa.Add((-5, 5)), iaa.Multiply((0.9, 1.1))])
List of augmenters:
* Add
* AddElementwise
* AdditiveGaussianNoise
* AdditiveLaplaceNoise
* AdditivePoissonNoise
* Multiply
* MultiplyElementwise
* Dropout
* CoarseDropout
* ReplaceElementwise
* ImpulseNoise
* SaltAndPepper
* CoarseSaltAndPepper
* Salt
* CoarseSalt
* Pepper
* CoarsePepper
* Invert
* ContrastNormalization
* JpegCompression
"""
from __future__ import print_function, division, absolute_import
from PIL import Image as PIL_Image
import imageio
import tempfile
import numpy as np
import cv2
from . import meta
from .. import imgaug as ia
from .. import parameters as iap
from .. import dtypes as iadt
class Add(meta.Augmenter):
"""
Add a value to all pixels in an image.
dtype support::
* ``uint8``: yes; fully tested
* ``uint16``: yes; tested
* ``uint32``: no
* ``uint64``: no
* ``int8``: yes; tested
* ``int16``: yes; tested
* ``int32``: no
* ``int64``: no
* ``float16``: yes; tested
* ``float32``: yes; tested
* ``float64``: no
* ``float128``: no
* ``bool``: yes; tested
Parameters
----------
value : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
Value to add to all pixels.
* If a number, then that value will be used for all images.
* If a tuple ``(a, b)``, then a value from the discrete range ``[a, b]``
will be used.
* If a list, then a random value will be sampled from that list per image.
* If a StochasticParameter, then a value will be sampled per image
from that parameter.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.Add(10)
always adds a value of 10 to all pixels in the image.
>>> aug = iaa.Add((-10, 10))
adds a value from the discrete range ``[-10 .. 10]`` to all pixels of
the input images. The exact value is sampled per image.
>>> aug = iaa.Add((-10, 10), per_channel=True)
adds a value from the discrete range ``[-10 .. 10]`` to all pixels of
the input images. The exact value is sampled per image AND channel,
i.e. to a red-channel it might add 5 while subtracting 7 from the
blue channel of the same image.
>>> aug = iaa.Add((-10, 10), per_channel=0.5)
same as previous example, but the `per_channel` feature is only active
for 50 percent of all images.
"""
def __init__(self, value=0, per_channel=False, name=None, deterministic=False, random_state=None):
super(Add, self).__init__(name=name, deterministic=deterministic, random_state=random_state)
self.value = iap.handle_continuous_param(value, "value", value_range=None, tuple_to_uniform=True,
list_to_choice=True)
self.per_channel = iap.handle_probability_param(per_channel, "per_channel")
def _augment_images(self, images, random_state, parents, hooks):
iadt.gate_dtypes(images,
allowed=["bool", "uint8", "uint16", "int8", "int16", "float16", "float32"],
disallowed=["uint32", "uint64", "uint128", "uint256", "int32", "int64", "int128", "int256",
"float64", "float96", "float128", "float256"],
augmenter=self)
input_dtypes = iadt.copy_dtypes_for_restore(images, force_list=True)
nb_images = len(images)
nb_channels_max = meta.estimate_max_number_of_channels(images)
rss = ia.derive_random_states(random_state, 2)
value_samples = self.value.draw_samples((nb_images, nb_channels_max), random_state=rss[0])
per_channel_samples = self.per_channel.draw_samples((nb_images,), random_state=rss[1])
gen = enumerate(zip(images, value_samples, per_channel_samples, input_dtypes))
for i, (image, value_samples_i, per_channel_samples_i, input_dtype) in gen:
nb_channels = image.shape[2]
# Example code to directly add images via image+sample (uint8 only)
# if per_channel_samples_i > 0.5:
# result = []
# image = image.astype(np.int16)
# value_samples_i = value_samples_i.astype(np.int16)
# for c, value in enumerate(value_samples_i[0:nb_channels]):
# result.append(np.clip(image[..., c:c+1] + value, 0, 255).astype(np.uint8))
# images[i] = np.concatenate(result, axis=2)
# else:
# images[i] = np.clip(
# image.astype(np.int16) + value_samples_i[0].astype(np.int16), 0, 255).astype(np.uint8)
if image.dtype.name == "uint8":
# Using this LUT approach is significantly faster than the else-block code (around 3-4x speedup)
# and is still faster than the simpler image+sample approach without LUT (about 10% at 64x64 and about
# 2x at 224x224 -- maybe dependent on installed BLAS libraries?)
value_samples_i = np.clip(np.round(value_samples_i), -255, 255).astype(np.int16)
value_range = np.arange(0, 256, dtype=np.int16)
if per_channel_samples_i > 0.5:
result = []
tables = np.tile(value_range[np.newaxis, :], (nb_channels, 1)) \
+ value_samples_i[0:nb_channels, np.newaxis]
tables = np.clip(tables, 0, 255).astype(image.dtype)
for c, table in enumerate(tables):
arr_aug = cv2.LUT(image[..., c], table)
result.append(arr_aug[..., np.newaxis])
images[i] = np.concatenate(result, axis=2)
else:
table = value_range + value_samples_i[0]
image_aug = cv2.LUT(image, np.clip(table, 0, 255).astype(image.dtype))
if image_aug.ndim == 2:
image_aug = image_aug[..., np.newaxis]
images[i] = image_aug
else:
if per_channel_samples_i > 0.5:
value = value_samples_i[0:nb_channels].reshape((1, 1, nb_channels))
else:
value = value_samples_i[0:1].reshape((1, 1, 1))
# We limit here the value range of the value parameter to the bytes in the image's dtype.
# This prevents overflow problems and makes it less likely that the image has to be up-casted, which
# again improves performance and saves memory. Note that this also enables more dtypes for image inputs.
# The downside is that the mul parameter is limited in its value range.
#
# We need 2* the itemsize of the image here to allow to shift the image's max value to the lowest
# possible value, e.g. for uint8 it must allow for -255 to 255.
itemsize = image.dtype.itemsize * 2
dtype_target = np.dtype("%s%d" % (value.dtype.kind, itemsize))
value = iadt.clip_to_dtype_value_range_(value, dtype_target, validate=True)
image, value = iadt.promote_array_dtypes_([image, value], dtypes=[image.dtype, dtype_target],
increase_itemsize_factor=2)
image = np.add(image, value, out=image, casting="no")
image = iadt.restore_dtypes_(image, input_dtype)
images[i] = image
return images
def _augment_heatmaps(self, heatmaps, random_state, parents, hooks):
return heatmaps
def _augment_keypoints(self, keypoints_on_images, random_state, parents, hooks):
return keypoints_on_images
def get_parameters(self):
return [self.value, self.per_channel]
# TODO merge this with Add
class AddElementwise(meta.Augmenter):
"""
Add values to the pixels of images with possibly different values for neighbouring pixels.
While the Add Augmenter adds a constant value per image, this one can
add different values (sampled per pixel).
dtype support::
* ``uint8``: yes; fully tested
* ``uint16``: yes; tested
* ``uint32``: no
* ``uint64``: no
* ``int8``: yes; tested
* ``int16``: yes; tested
* ``int32``: no
* ``int64``: no
* ``float16``: yes; tested
* ``float32``: yes; tested
* ``float64``: no
* ``float128``: no
* ``bool``: yes; tested
Parameters
----------
value : int or tuple of int or list of int or imgaug.parameters.StochasticParameter, optional
Value to add to the pixels.
* If an int, then that value will be used for all images.
* If a tuple ``(a, b)``, then values from the discrete range ``[a .. b]``
will be sampled.
* If a list of integers, a random value will be sampled from the list
per image.
* If a StochasticParameter, then values will be sampled per pixel
(and possibly channel) from that parameter.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.AddElementwise(10)
always adds a value of 10 to all pixels in the image.
>>> aug = iaa.AddElementwise((-10, 10))
samples per pixel a value from the discrete range ``[-10 .. 10]`` and
adds that value to the pixel.
>>> aug = iaa.AddElementwise((-10, 10), per_channel=True)
samples per pixel *and channel* a value from the discrete
range ``[-10 .. 10]`` ands adds it to the pixel's value. Therefore,
added values may differ between channels of the same pixel.
>>> aug = iaa.AddElementwise((-10, 10), per_channel=0.5)
same as previous example, but the `per_channel` feature is only active
for 50 percent of all images.
"""
def __init__(self, value=0, per_channel=False, name=None, deterministic=False, random_state=None):
super(AddElementwise, self).__init__(name=name, deterministic=deterministic, random_state=random_state)
# TODO open to continous, similar to Add
self.value = iap.handle_discrete_param(value, "value", value_range=(-255, 255), tuple_to_uniform=True,
list_to_choice=True, allow_floats=False)
self.per_channel = iap.handle_probability_param(per_channel, "per_channel")
def _augment_images(self, images, random_state, parents, hooks):
iadt.gate_dtypes(images,
allowed=["bool", "uint8", "uint16", "int8", "int16", "float16", "float32"],
disallowed=["uint32", "uint64", "uint128", "uint256", "int32", "int64", "int128", "int256",
"float64", "float96", "float128", "float256"],
augmenter=self)
input_dtypes = iadt.copy_dtypes_for_restore(images, force_list=True)
nb_images = len(images)
rss = ia.derive_random_states(random_state, nb_images+1)
per_channel_samples = self.per_channel.draw_samples((nb_images,), random_state=rss[-1])
gen = enumerate(zip(images, per_channel_samples, rss[:-1], input_dtypes))
for i, (image, per_channel_samples_i, rs, input_dtype) in gen:
height, width, nb_channels = image.shape
sample_shape = (height, width, nb_channels if per_channel_samples_i > 0.5 else 1)
value = self.value.draw_samples(sample_shape, random_state=rs)
if image.dtype.name == "uint8":
# This special uint8 block is around 60-100% faster than the else-block further below (more speedup
# for smaller images).
#
# Also tested to instead compute min/max of image and value and then only convert image/value dtype
# if actually necessary, but that was like 20-30% slower, even for 224x224 images.
#
if value.dtype.kind == "f":
value = np.round(value)
image = image.astype(np.int16)
value = np.clip(value, -255, 255).astype(np.int16)
image_aug = image + value
image_aug = np.clip(image_aug, 0, 255).astype(np.uint8)
images[i] = image_aug
else:
# We limit here the value range of the value parameter to the bytes in the image's dtype.
# This prevents overflow problems and makes it less likely that the image has to be up-casted, which
# again improves performance and saves memory. Note that this also enables more dtypes for image inputs.
# The downside is that the mul parameter is limited in its value range.
#
# We need 2* the itemsize of the image here to allow to shift the image's max value to the lowest
# possible value, e.g. for uint8 it must allow for -255 to 255.
itemsize = image.dtype.itemsize * 2
dtype_target = np.dtype("%s%d" % (value.dtype.kind, itemsize))
value = iadt.clip_to_dtype_value_range_(value, dtype_target, validate=100)
if value.shape[2] == 1:
value = np.tile(value, (1, 1, nb_channels))
image, value = iadt.promote_array_dtypes_([image, value], dtypes=[image.dtype, dtype_target],
increase_itemsize_factor=2)
image = np.add(image, value, out=image, casting="no")
image = iadt.restore_dtypes_(image, input_dtype)
images[i] = image
return images
def _augment_heatmaps(self, heatmaps, random_state, parents, hooks):
return heatmaps
def _augment_keypoints(self, keypoints_on_images, random_state, parents, hooks):
return keypoints_on_images
def get_parameters(self):
return [self.value, self.per_channel]
def AdditiveGaussianNoise(loc=0, scale=0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Add gaussian noise (aka white noise) to images.
dtype support::
See ``imgaug.augmenters.arithmetic.AddElementwise``.
Parameters
----------
loc : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
Mean of the normal distribution that generates the noise.
* If a number, exactly that value will be used.
* If a tuple ``(a, b)``, a random value from the range ``a <= x <= b`` will
be sampled per image.
* If a list, then a random value will be sampled from that list per
image.
* If a StochasticParameter, a value will be sampled from the
parameter per image.
scale : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
Standard deviation of the normal distribution that generates the noise.
Must be ``>= 0``. If 0 then only `loc` will be used.
* If an int or float, exactly that value will be used.
* If a tuple ``(a, b)``, a random value from the range ``a <= x <= b`` will
be sampled per image.
* If a list, then a random value will be sampled from that list per
image.
* If a StochasticParameter, a value will be sampled from the
parameter per image.
per_channel : bool or float, optional
Whether to use the same noise value per pixel for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.AdditiveGaussianNoise(scale=0.1*255)
adds gaussian noise from the distribution ``N(0, 0.1*255)`` to images.
>>> aug = iaa.AdditiveGaussianNoise(scale=(0, 0.1*255))
adds gaussian noise from the distribution ``N(0, s)`` to images,
where s is sampled per image from the range ``0 <= s <= 0.1*255``.
>>> aug = iaa.AdditiveGaussianNoise(scale=0.1*255, per_channel=True)
adds gaussian noise from the distribution ``N(0, 0.1*255)`` to images,
where the noise value is different per pixel *and* channel (e.g. a
different one for red, green and blue channels for the same pixel).
>>> aug = iaa.AdditiveGaussianNoise(scale=0.1*255, per_channel=0.5)
adds gaussian noise from the distribution ``N(0, 0.1*255)`` to images,
where the noise value is sometimes (50 percent of all cases) the same
per pixel for all channels and sometimes different (other 50 percent).
"""
loc2 = iap.handle_continuous_param(loc, "loc", value_range=None, tuple_to_uniform=True, list_to_choice=True)
scale2 = iap.handle_continuous_param(scale, "scale", value_range=(0, None), tuple_to_uniform=True,
list_to_choice=True)
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return AddElementwise(iap.Normal(loc=loc2, scale=scale2), per_channel=per_channel, name=name,
deterministic=deterministic, random_state=random_state)
def AdditiveLaplaceNoise(loc=0, scale=0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Add laplace noise to images.
The laplace distribution is similar to the gaussian distribution, but has puts weight on the long tail.
Hence, this noise will add more outliers (very high/low values). It is somewhere between gaussian noise and
salt and pepper noise.
Values of around ``255 * 0.05`` for `scale` lead to visible noise (for uint8).
Values of around ``255 * 0.10`` for `scale` lead to very visible noise (for uint8).
It is recommended to usually set `per_channel` to True.
dtype support::
See ``imgaug.augmenters.arithmetic.AddElementwise``.
Parameters
----------
loc : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
Mean of the laplace distribution that generates the noise.
* If a number, exactly that value will be used.
* If a tuple ``(a, b)``, a random value from the range ``a <= x <= b`` will
be sampled per image.
* If a list, then a random value will be sampled from that list per
image.
* If a StochasticParameter, a value will be sampled from the
parameter per image.
scale : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
Standard deviation of the laplace distribution that generates the noise.
Must be ``>= 0``. If 0 then only `loc` will be used.
Recommended to be around ``255 * 0.05``.
* If an int or float, exactly that value will be used.
* If a tuple ``(a, b)``, a random value from the range ``a <= x <= b`` will
be sampled per image.
* If a list, then a random value will be sampled from that list per
image.
* If a StochasticParameter, a value will be sampled from the
parameter per image.
per_channel : bool or float, optional
Whether to use the same noise value per pixel for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.AdditiveLaplaceNoise(scale=0.1*255)
Adds laplace noise from the distribution ``Laplace(0, 0.1*255)`` to images.
>>> aug = iaa.AdditiveLaplaceNoise(scale=(0, 0.1*255))
Adds laplace noise from the distribution ``Laplace(0, s)`` to images,
where s is sampled per image from the range ``0 <= s <= 0.1*255``.
>>> aug = iaa.AdditiveLaplaceNoise(scale=0.1*255, per_channel=True)
Adds laplace noise from the distribution ``Laplace(0, 0.1*255)`` to images,
where the noise value is different per pixel *and* channel (e.g. a
different one for red, green and blue channels for the same pixel).
>>> aug = iaa.AdditiveLaplaceNoise(scale=0.1*255, per_channel=0.5)
Adds laplace noise from the distribution ``Laplace(0, 0.1*255)`` to images,
where the noise value is sometimes (50 percent of all cases) the same
per pixel for all channels and sometimes different (other 50 percent).
"""
loc2 = iap.handle_continuous_param(loc, "loc", value_range=None, tuple_to_uniform=True, list_to_choice=True)
scale2 = iap.handle_continuous_param(scale, "scale", value_range=(0, None), tuple_to_uniform=True,
list_to_choice=True)
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return AddElementwise(iap.Laplace(loc=loc2, scale=scale2), per_channel=per_channel, name=name,
deterministic=deterministic, random_state=random_state)
def AdditivePoissonNoise(lam=0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Create an augmenter to add poisson noise to images.
Poisson noise is comparable to gaussian noise as in ``AdditiveGaussianNoise``, but the values are sampled from
a poisson distribution instead of a gaussian distribution. As poisson distributions produce only positive numbers,
the sign of the sampled values are here randomly flipped.
Values of around ``10.0`` for `lam` lead to visible noise (for uint8).
Values of around ``20.0`` for `lam` lead to very visible noise (for uint8).
It is recommended to usually set `per_channel` to True.
dtype support::
See ``imgaug.augmenters.arithmetic.AddElementwise``.
Parameters
----------
lam : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
Lambda parameter of the poisson distribution. Recommended values are around ``0.0`` to ``10.0``.
* If a number, exactly that value will be used.
* If a tuple ``(a, b)``, a random value from the range ``a <= x <= b`` will
be sampled per image.
* If a list, then a random value will be sampled from that list per image.
* If a StochasticParameter, a value will be sampled from the
parameter per image.
per_channel : bool or float, optional
Whether to use the same noise value per pixel for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.AdditivePoissonNoise(lam=5.0)
Adds poisson noise sampled from ``Poisson(5.0)`` to images.
>>> aug = iaa.AdditivePoissonNoise(lam=(0.0, 10.0))
Adds poisson noise sampled from ``Poisson(x)`` to images, where ``x`` is randomly sampled per image from the
interval ``[0.0, 10.0]``.
>>> aug = iaa.AdditivePoissonNoise(lam=5.0, per_channel=True)
Adds poisson noise sampled from ``Poisson(5.0)`` to images,
where the values are different per pixel *and* channel (e.g. a
different one for red, green and blue channels for the same pixel).
>>> aug = iaa.AdditivePoissonNoise(lam=(0.0, 10.0), per_channel=True)
Adds poisson noise sampled from ``Poisson(x)`` to images,
with ``x`` being sampled from ``uniform(0.0, 10.0)`` per image, pixel and channel.
This is the *recommended* configuration.
>>> aug = iaa.AdditivePoissonNoise(lam=2, per_channel=0.5)
Adds poisson noise sampled from the distribution ``Poisson(2)`` to images,
where the values are sometimes (50 percent of all cases) the same
per pixel for all channels and sometimes different (other 50 percent).
"""
lam2 = iap.handle_continuous_param(lam, "lam", value_range=(0, None), tuple_to_uniform=True,
list_to_choice=True)
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return AddElementwise(iap.RandomSign(iap.Poisson(lam=lam2)), per_channel=per_channel, name=name,
deterministic=deterministic, random_state=random_state)
class Multiply(meta.Augmenter):
"""
Multiply all pixels in an image with a specific value.
This augmenter can be used to make images lighter or darker.
dtype support::
* ``uint8``: yes; fully tested
* ``uint16``: yes; tested
* ``uint32``: no
* ``uint64``: no
* ``int8``: yes; tested
* ``int16``: yes; tested
* ``int32``: no
* ``int64``: no
* ``float16``: yes; tested
* ``float32``: yes; tested
* ``float64``: no
* ``float128``: no
* ``bool``: yes; tested
Note: tests were only conducted for rather small multipliers, around -10.0 to +10.0.
In general, the multipliers sampled from `mul` must be in a value range that corresponds to
the input image's dtype. E.g. if the input image has dtype uint16 and the samples generated
from `mul` are float64, this augmenter will still force all samples to be within the value
range of float16, as it has the same number of bytes (two) as uint16. This is done to
make overflows less likely to occur.
Parameters
----------
mul : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
The value with which to multiply the pixel values in each image.
* If a number, then that value will always be used.
* If a tuple ``(a, b)``, then a value from the range ``a <= x <= b`` will
be sampled per image and used for all pixels.
* If a list, then a random value will be sampled from that list per image.
* If a StochasticParameter, then that parameter will be used to
sample a new value per image.
per_channel : bool or float, optional
Whether to use the same multiplier per pixel for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.Multiply(2.0)
would multiply all images by a factor of 2, making the images
significantly brighter.
>>> aug = iaa.Multiply((0.5, 1.5))
would multiply images by a random value from the range ``0.5 <= x <= 1.5``,
making some images darker and others brighter.
"""
def __init__(self, mul=1.0, per_channel=False, name=None, deterministic=False, random_state=None):
super(Multiply, self).__init__(name=name, deterministic=deterministic, random_state=random_state)
self.mul = iap.handle_continuous_param(mul, "mul", value_range=None, tuple_to_uniform=True,
list_to_choice=True)
self.per_channel = iap.handle_probability_param(per_channel, "per_channel")
def _augment_images(self, images, random_state, parents, hooks):
iadt.gate_dtypes(images,
allowed=["bool", "uint8", "uint16", "int8", "int16", "float16", "float32"],
disallowed=["uint32", "uint64", "uint128", "uint256", "int32", "int64", "int128", "int256",
"float64", "float96", "float128", "float256"],
augmenter=self)
input_dtypes = iadt.copy_dtypes_for_restore(images, force_list=True)
nb_images = len(images)
nb_channels_max = meta.estimate_max_number_of_channels(images)
rss = ia.derive_random_states(random_state, 2)
mul_samples = self.mul.draw_samples((nb_images, nb_channels_max), random_state=rss[0])
per_channel_samples = self.per_channel.draw_samples((nb_images,), random_state=rss[1])
gen = enumerate(zip(images, mul_samples, per_channel_samples, input_dtypes))
for i, (image, mul_samples_i, per_channel_samples_i, input_dtype) in gen:
nb_channels = image.shape[2]
# Example code to directly multiply images via image*sample (uint8 only) -- apparently slower than LUT
# if per_channel_samples_i > 0.5:
# result = []
# image = image.astype(np.float32)
# mul_samples_i = mul_samples_i.astype(np.float32)
# for c, mul in enumerate(mul_samples_i[0:nb_channels]):
# result.append(np.clip(image[..., c:c+1] * mul, 0, 255).astype(np.uint8))
# images[i] = np.concatenate(result, axis=2)
# else:
# images[i] = np.clip(
# image.astype(np.float32) * mul_samples_i[0].astype(np.float32), 0, 255).astype(np.uint8)
if image.dtype.name == "uint8":
# Using this LUT approach is significantly faster than else-block code (more than 10x speedup)
# and is still faster than the simpler image*sample approach without LUT (1.5-3x speedup,
# maybe dependent on installed BLAS libraries?)
value_range = np.arange(0, 256, dtype=np.float32)
if per_channel_samples_i > 0.5:
result = []
mul_samples_i = mul_samples_i.astype(np.float32)
tables = np.tile(value_range[np.newaxis, :], (nb_channels, 1)) \
* mul_samples_i[0:nb_channels, np.newaxis]
tables = np.clip(tables, 0, 255).astype(image.dtype)
for c, table in enumerate(tables):
arr_aug = cv2.LUT(image[..., c], table)
result.append(arr_aug[..., np.newaxis])
images[i] = np.concatenate(result, axis=2)
else:
table = value_range * mul_samples_i[0].astype(np.float32)
image_aug = cv2.LUT(image, np.clip(table, 0, 255).astype(image.dtype))
if image_aug.ndim == 2:
image_aug = image_aug[..., np.newaxis]
images[i] = image_aug
else:
# TODO estimate via image min/max values whether a resolution increase is necessary
if per_channel_samples_i > 0.5:
mul = mul_samples_i[0:nb_channels].reshape((1, 1, nb_channels))
else:
mul = mul_samples_i[0:1].reshape((1, 1, 1))
mul_min = np.min(mul)
mul_max = np.max(mul)
is_not_increasing_value_range = (-1 <= mul_min <= 1) and (-1 <= mul_max <= 1)
# We limit here the value range of the mul parameter to the bytes in the image's dtype.
# This prevents overflow problems and makes it less likely that the image has to be up-casted, which
# again improves performance and saves memory. Note that this also enables more dtypes for image inputs.
# The downside is that the mul parameter is limited in its value range.
itemsize = max(image.dtype.itemsize, 2 if mul.dtype.kind == "f" else 1) # float min itemsize is 2 not 1
dtype_target = np.dtype("%s%d" % (mul.dtype.kind, itemsize))
mul = iadt.clip_to_dtype_value_range_(mul, dtype_target, validate=True)
image, mul = iadt.promote_array_dtypes_(
[image, mul],
dtypes=[image.dtype, dtype_target],
increase_itemsize_factor=1 if is_not_increasing_value_range else 2)
image = np.multiply(image, mul, out=image, casting="no")
image = iadt.restore_dtypes_(image, input_dtype)
images[i] = image
return images
def _augment_heatmaps(self, heatmaps, random_state, parents, hooks):
return heatmaps
def _augment_keypoints(self, keypoints_on_images, random_state, parents, hooks):
return keypoints_on_images
def get_parameters(self):
return [self.mul, self.per_channel]
# TODO merge with Multiply
class MultiplyElementwise(meta.Augmenter):
"""
Multiply values of pixels with possibly different values for neighbouring pixels.
While the Multiply Augmenter uses a constant multiplier per image,
this one can use different multipliers per pixel.
dtype support::
* ``uint8``: yes; fully tested
* ``uint16``: yes; tested
* ``uint32``: no
* ``uint64``: no
* ``int8``: yes; tested
* ``int16``: yes; tested
* ``int32``: no
* ``int64``: no
* ``float16``: yes; tested
* ``float32``: yes; tested
* ``float64``: no
* ``float128``: no
* ``bool``: yes; tested
Note: tests were only conducted for rather small multipliers, around -10.0 to +10.0.
In general, the multipliers sampled from `mul` must be in a value range that corresponds to
the input image's dtype. E.g. if the input image has dtype uint16 and the samples generated
from `mul` are float64, this augmenter will still force all samples to be within the value
range of float16, as it has the same number of bytes (two) as uint16. This is done to
make overflows less likely to occur.
Parameters
----------
mul : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
The value by which to multiply the pixel values in the image.
* If a number, then that value will always be used.
* If a tuple ``(a, b)``, then a value from the range ``a <= x <= b`` will
be sampled per image and pixel.
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then that parameter will be used to
sample a new value per image and pixel.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> from imgaug import augmenters as iaa
>>> aug = iaa.MultiplyElementwise(2.0)
multiply all images by a factor of 2.0, making them significantly
bighter.
>>> aug = iaa.MultiplyElementwise((0.5, 1.5))
samples per pixel a value from the range ``0.5 <= x <= 1.5`` and
multiplies the pixel with that value.
>>> aug = iaa.MultiplyElementwise((0.5, 1.5), per_channel=True)
samples per pixel *and channel* a value from the range
``0.5 <= x <= 1.5`` ands multiplies the pixel by that value. Therefore,
added multipliers may differ between channels of the same pixel.
>>> aug = iaa.MultiplyElementwise((0.5, 1.5), per_channel=0.5)
same as previous example, but the `per_channel` feature is only active
for 50 percent of all images.
"""
def __init__(self, mul=1.0, per_channel=False, name=None, deterministic=False, random_state=None):
super(MultiplyElementwise, self).__init__(name=name, deterministic=deterministic, random_state=random_state)
self.mul = iap.handle_continuous_param(mul, "mul", value_range=None, tuple_to_uniform=True,
list_to_choice=True)
self.per_channel = iap.handle_probability_param(per_channel, "per_channel")
def _augment_images(self, images, random_state, parents, hooks):
iadt.gate_dtypes(images,
allowed=["bool", "uint8", "uint16", "int8", "int16", "float16", "float32"],
disallowed=["uint32", "uint64", "uint128", "uint256", "int32", "int64", "int128", "int256",
"float64", "float96", "float128", "float256"],
augmenter=self)
input_dtypes = iadt.copy_dtypes_for_restore(images, force_list=True)
nb_images = len(images)
rss = ia.derive_random_states(random_state, nb_images+1)
per_channel_samples = self.per_channel.draw_samples((nb_images,), random_state=rss[-1])
is_mul_binomial = isinstance(self.mul, iap.Binomial) or (
isinstance(self.mul, iap.FromLowerResolution) and isinstance(self.mul.other_param, iap.Binomial)
)
gen = enumerate(zip(images, per_channel_samples, rss[:-1], input_dtypes))
for i, (image, per_channel_samples_i, rs, input_dtype) in gen:
height, width, nb_channels = image.shape
sample_shape = (height, width, nb_channels if per_channel_samples_i > 0.5 else 1)
mul = self.mul.draw_samples(sample_shape, random_state=rs)
# TODO let Binomial return boolean mask directly instead of [0, 1] integers?
# hack to improve performance for Dropout and CoarseDropout
# converts mul samples to mask if mul is binomial
if mul.dtype.kind != "b" and is_mul_binomial:
mul = mul.astype(bool, copy=False)
if mul.dtype.kind == "b":
images[i] *= mul
elif image.dtype.name == "uint8":
# This special uint8 block is around 60-100% faster than the else-block further below (more speedup
# for larger images).
#
if mul.dtype.kind == "f":
# interestingly, float32 is here significantly faster than float16
# TODO is that system dependent?
# TODO does that affect int8-int32 too?
mul = mul.astype(np.float32, copy=False)
image_aug = image.astype(np.float32)
else:
mul = mul.astype(np.int16, copy=False)
image_aug = image.astype(np.int16)
image_aug = np.multiply(image_aug, mul, casting="no", out=image_aug)
images[i] = iadt.restore_dtypes_(image_aug, np.uint8, round=False)
else:
# TODO maybe introduce to stochastic parameters some way to get the possible min/max values,
# could make things faster for dropout to get 0/1 min/max from the binomial
mul_min = np.min(mul)
mul_max = np.max(mul)
is_not_increasing_value_range = (-1 <= mul_min <= 1) and (-1 <= mul_max <= 1)
# We limit here the value range of the mul parameter to the bytes in the image's dtype.
# This prevents overflow problems and makes it less likely that the image has to be up-casted, which
# again improves performance and saves memory. Note that this also enables more dtypes for image inputs.
# The downside is that the mul parameter is limited in its value range.
itemsize = max(image.dtype.itemsize, 2 if mul.dtype.kind == "f" else 1) # float min itemsize is 2
dtype_target = np.dtype("%s%d" % (mul.dtype.kind, itemsize))
mul = iadt.clip_to_dtype_value_range_(mul, dtype_target, validate=True,
validate_values=(mul_min, mul_max))
if mul.shape[2] == 1:
mul = np.tile(mul, (1, 1, nb_channels))
image, mul = iadt.promote_array_dtypes_(
[image, mul],
dtypes=[image, dtype_target],
increase_itemsize_factor=1 if is_not_increasing_value_range else 2)
image = np.multiply(image, mul, out=image, casting="no")
image = iadt.restore_dtypes_(image, input_dtype)
images[i] = image
return images
def _augment_heatmaps(self, heatmaps, random_state, parents, hooks):
return heatmaps
def _augment_keypoints(self, keypoints_on_images, random_state, parents, hooks):
return keypoints_on_images
def get_parameters(self):
return [self.mul, self.per_channel]
def Dropout(p=0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Augmenter that sets a certain fraction of pixels in images to zero.
dtype support::
See ``imgaug.augmenters.arithmetic.MultiplyElementwise``.
Parameters
----------
p : float or tuple of float or imgaug.parameters.StochasticParameter, optional
The probability of any pixel being dropped (i.e. set to zero).
* If a float, then that value will be used for all images. A value
of 1.0 would mean that all pixels will be dropped and 0.0 that
no pixels would be dropped. A value of 0.05 corresponds to 5
percent of all pixels dropped.
* If a tuple ``(a, b)``, then a value p will be sampled from the
range ``a <= p <= b`` per image and be used as the pixel's dropout
probability.
* If a StochasticParameter, then this parameter will be used to
determine per pixel whether it should be dropped (sampled value
of 0) or shouldn't (sampled value of 1).
If you instead want to provide the probability as a stochastic
parameter, you can usually do ``imgaug.parameters.Binomial(1-p)``
to convert parameter `p` to a 0/1 representation.
per_channel : bool or float, optional
Whether to use the same value (is dropped / is not dropped)
for all channels of a pixel (False) or to sample a new value for each
channel (True).
If this value is a float p, then for p percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.Dropout(0.02)
drops 2 percent of all pixels.
>>> aug = iaa.Dropout((0.0, 0.05))
drops in each image a random fraction of all pixels, where the fraction
is in the range ``0.0 <= x <= 0.05``.
>>> aug = iaa.Dropout(0.02, per_channel=True)
drops 2 percent of all pixels in a channel-wise fashion, i.e. it is unlikely
for any pixel to have all channels set to zero (black pixels).
>>> aug = iaa.Dropout(0.02, per_channel=0.5)
same as previous example, but the `per_channel` feature is only active
for 50 percent of all images.
"""
if ia.is_single_number(p):
p2 = iap.Binomial(1 - p)
elif ia.is_iterable(p):
ia.do_assert(len(p) == 2)
ia.do_assert(p[0] < p[1])
ia.do_assert(0 <= p[0] <= 1.0)
ia.do_assert(0 <= p[1] <= 1.0)
p2 = iap.Binomial(iap.Uniform(1 - p[1], 1 - p[0]))
elif isinstance(p, iap.StochasticParameter):
p2 = p
else:
raise Exception("Expected p to be float or int or StochasticParameter, got %s." % (type(p),))
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return MultiplyElementwise(p2, per_channel=per_channel, name=name, deterministic=deterministic,
random_state=random_state)
def CoarseDropout(p=0, size_px=None, size_percent=None, per_channel=False, min_size=4, name=None, deterministic=False,
random_state=None):
"""
Augmenter that sets rectangular areas within images to zero.
In contrast to Dropout, these areas can have larger sizes.
(E.g. you might end up with three large black rectangles in an image.)
Note that the current implementation leads to correlated sizes,
so when there is one large area that is dropped, there is a high likelihood
that all other dropped areas are also large.
This method is implemented by generating the dropout mask at a
lower resolution (than the image has) and then upsampling the mask
before dropping the pixels.
dtype support::
See ``imgaug.augmenters.arithmetic.MultiplyElementwise``.
Parameters
----------
p : float or tuple of float or imgaug.parameters.StochasticParameter, optional
The probability of any pixel being dropped (i.e. set to zero).
* If a float, then that value will be used for all pixels. A value
of 1.0 would mean, that all pixels will be dropped. A value of
0.0 would lead to no pixels being dropped.
* If a tuple ``(a, b)``, then a value p will be sampled from the
range ``a <= p <= b`` per image and be used as the pixel's dropout
probability.
* If a StochasticParameter, then this parameter will be used to
determine per pixel whether it should be dropped (sampled value
of 0) or shouldn't (sampled value of 1).
size_px : int or tuple of int or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the dropout
mask in absolute pixel dimensions.
* If an integer, then that size will be used for both height and
width. E.g. a value of 3 would lead to a ``3x3`` mask, which is then
upsampled to ``HxW``, where ``H`` is the image size and W the image width.
* If a tuple ``(a, b)``, then two values ``M``, ``N`` will be sampled from the
range ``[a..b]`` and the mask will be generated at size ``MxN``, then
upsampled to ``HxW``.
* If a StochasticParameter, then this parameter will be used to
determine the sizes. It is expected to be discrete.
size_percent : float or tuple of float or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the dropout
mask *in percent* of the input image.
* If a float, then that value will be used as the percentage of the
height and width (relative to the original size). E.g. for value
p, the mask will be sampled from ``(p*H)x(p*W)`` and later upsampled
to ``HxW``.
* If a tuple ``(a, b)``, then two values ``m``, ``n`` will be sampled from the
interval ``(a, b)`` and used as the percentages, i.e the mask size
will be ``(m*H)x(n*W)``.
* If a StochasticParameter, then this parameter will be used to
sample the percentage values. It is expected to be continuous.
per_channel : bool or float, optional
Whether to use the same value (is dropped / is not dropped)
for all channels of a pixel (False) or to sample a new value for each
channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
min_size : int, optional
Minimum size of the low resolution mask, both width and height. If
`size_percent` or `size_px` leads to a lower value than this, `min_size`
will be used instead. This should never have a value of less than 2,
otherwise one may end up with a ``1x1`` low resolution mask, leading easily
to the whole image being dropped.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.CoarseDropout(0.02, size_percent=0.5)
drops 2 percent of all pixels on an lower-resolution image that has
50 percent of the original image's size, leading to dropped areas that
have roughly 2x2 pixels size.
>>> aug = iaa.CoarseDropout((0.0, 0.05), size_percent=(0.05, 0.5))
generates a dropout mask at 5 to 50 percent of image's size. In that mask,
0 to 5 percent of all pixels are dropped (random per image).
>>> aug = iaa.CoarseDropout((0.0, 0.05), size_px=(2, 16))
same as previous example, but the lower resolution image has 2 to 16 pixels
size.
>>> aug = iaa.CoarseDropout(0.02, size_percent=0.5, per_channel=True)
drops 2 percent of all pixels at 50 percent resolution (2x2 sizes)
in a channel-wise fashion, i.e. it is unlikely
for any pixel to have all channels set to zero (black pixels).
>>> aug = iaa.CoarseDropout(0.02, size_percent=0.5, per_channel=0.5)
same as previous example, but the `per_channel` feature is only active
for 50 percent of all images.
"""
if ia.is_single_number(p):
p2 = iap.Binomial(1 - p)
elif ia.is_iterable(p):
ia.do_assert(len(p) == 2)
ia.do_assert(p[0] < p[1])
ia.do_assert(0 <= p[0] <= 1.0)
ia.do_assert(0 <= p[1] <= 1.0)
p2 = iap.Binomial(iap.Uniform(1 - p[1], 1 - p[0]))
elif isinstance(p, iap.StochasticParameter):
p2 = p
else:
raise Exception("Expected p to be float or int or StochasticParameter, got %s." % (type(p),))
if size_px is not None:
p3 = iap.FromLowerResolution(other_param=p2, size_px=size_px, min_size=min_size)
elif size_percent is not None:
p3 = iap.FromLowerResolution(other_param=p2, size_percent=size_percent, min_size=min_size)
else:
raise Exception("Either size_px or size_percent must be set.")
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return MultiplyElementwise(p3, per_channel=per_channel, name=name, deterministic=deterministic,
random_state=random_state)
class ReplaceElementwise(meta.Augmenter):
"""
Replace pixels in an image with new values.
dtype support::
* ``uint8``: yes; fully tested
* ``uint16``: yes; tested
* ``uint32``: yes; tested
* ``uint64``: no (1)
* ``int8``: yes; tested
* ``int16``: yes; tested
* ``int32``: yes; tested
* ``int64``: yes; tested
* ``float16``: yes; tested
* ``float32``: yes; tested
* ``float64``: yes; tested
* ``float128``: no
* ``bool``: yes; tested
- (1) uint64 is currently not supported, because iadt.clip_to_dtype_value_range_() does not
support it, which again is because numpy.clip() seems to not support it.
Parameters
----------
mask : float or tuple of float or list of float or imgaug.parameters.StochasticParameter
Mask that indicates the pixels that are supposed to be replaced.
The mask will be thresholded with 0.5. A value of 1 then indicates a
pixel that is supposed to be replaced.
* If this is a float, then that value will be used as the
probability of being a 1 per pixel.
* If a tuple ``(a, b)``, then the probability will be sampled per image
from the range ``a <= x <= b``.
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then this parameter will be used to
sample a mask.
replacement : number or tuple of number or list of number or imgaug.parameters.StochasticParameter
The replacement to use at all locations that are marked as `1` in the mask.
* If this is a number, then that value will always be used as the
replacement.
* If a tuple ``(a, b)``, then the replacement will be sampled pixelwise
from the range ``a <= x <= b``.
* If a list of number, then a random value will be picked from
that list as the replacement per pixel.
* If a StochasticParameter, then this parameter will be used sample
pixelwise replacement values.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = ReplaceElementwise(0.05, [0, 255])
Replace 5 percent of all pixels in each image by either 0 or 255.
"""
def __init__(self, mask, replacement, per_channel=False, name=None, deterministic=False, random_state=None):
super(ReplaceElementwise, self).__init__(name=name, deterministic=deterministic, random_state=random_state)
self.mask = iap.handle_probability_param(mask, "mask", tuple_to_uniform=True, list_to_choice=True)
self.replacement = iap.handle_continuous_param(replacement, "replacement")
self.per_channel = iap.handle_probability_param(per_channel, "per_channel")
def _augment_images(self, images, random_state, parents, hooks):
iadt.gate_dtypes(images,
allowed=["bool", "uint8", "uint16", "uint32", "int8", "int16", "int32", "int64",
"float16", "float32", "float64"],
disallowed=["uint64", "uint128", "uint256", "int64", "int128", "int256",
"float96", "float128", "float256"],
augmenter=self)
nb_images = len(images)
rss = ia.derive_random_states(random_state, 2*nb_images+1)
per_channel_samples = self.per_channel.draw_samples((nb_images,), random_state=rss[-1])
gen = zip(images, per_channel_samples, rss[:-1:2], rss[1:-1:2])
for image, per_channel_i, rs_mask, rs_replacement in gen:
height, width, nb_channels = image.shape
sampling_shape = (height, width, nb_channels if per_channel_i > 0.5 else 1)
mask_samples = self.mask.draw_samples(sampling_shape, random_state=rs_mask)
# This is slightly faster (~20%) for masks that are True at many locations, but slower (~50%) for masks
# with few Trues, which is probably the more common use-case:
# replacement_samples = self.replacement.draw_samples(sampling_shape, random_state=rs_replacement)
#
# # round, this makes 0.2 e.g. become 0 in case of boolean image (otherwise replacing values with 0.2 would
# # lead to True instead of False).
# if image.dtype.kind in ["i", "u", "b"] and replacement_samples.dtype.kind == "f":
# replacement_samples = np.round(replacement_samples)
#
# replacement_samples = iadt.clip_to_dtype_value_range_(replacement_samples, image.dtype, validate=False)
# replacement_samples = replacement_samples.astype(image.dtype, copy=False)
#
# if sampling_shape[2] == 1:
# mask_samples = np.tile(mask_samples, (1, 1, nb_channels))
# replacement_samples = np.tile(replacement_samples, (1, 1, nb_channels))
# mask_thresh = mask_samples > 0.5
# image[mask_thresh] = replacement_samples[mask_thresh]
if sampling_shape[2] == 1:
mask_samples = np.tile(mask_samples, (1, 1, nb_channels))
mask_thresh = mask_samples > 0.5
replacement_samples = self.replacement.draw_samples((int(np.sum(mask_thresh)),),
random_state=rs_replacement)
# round, this makes 0.2 e.g. become 0 in case of boolean image (otherwise replacing values with 0.2 would
# lead to True instead of False).
if image.dtype.kind in ["i", "u", "b"] and replacement_samples.dtype.kind == "f":
replacement_samples = np.round(replacement_samples)
replacement_samples = iadt.clip_to_dtype_value_range_(replacement_samples, image.dtype, validate=False)
replacement_samples = replacement_samples.astype(image.dtype, copy=False)
image[mask_thresh] = replacement_samples
return images
def _augment_heatmaps(self, heatmaps, random_state, parents, hooks):
return heatmaps
def _augment_keypoints(self, keypoints_on_images, random_state, parents, hooks):
return keypoints_on_images
def get_parameters(self):
return [self.mask, self.replacement, self.per_channel]
def ImpulseNoise(p=0, name=None, deterministic=False, random_state=None):
"""
Creates an augmenter to apply impulse noise to an image.
This is identical to ``SaltAndPepper``, except that per_channel is always set to True.
dtype support::
See ``imgaug.augmenters.arithmetic.SaltAndPepper``.
"""
return SaltAndPepper(p=p, per_channel=True, name=name, deterministic=deterministic, random_state=random_state)
def SaltAndPepper(p=0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Adds salt and pepper noise to an image, i.e. some white-ish and black-ish pixels.
dtype support::
See ``imgaug.augmenters.arithmetic.ReplaceElementwise``.
Parameters
----------
p : float or tuple of float or list of float or imgaug.parameters.StochasticParameter, optional
Probability of changing a pixel to salt/pepper noise.
* If a float, then that value will be used for all images as the
probability.
* If a tuple ``(a, b)``, then a probability will be sampled per image
from the range ``a <= x <= b``.
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then this parameter will be used as
the *mask*, i.e. it is expected to contain values between
0.0 and 1.0, where 1.0 means that salt/pepper is to be added
at that location.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.SaltAndPepper(0.05)
Replaces 5 percent of all pixels with salt/pepper.
"""
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return ReplaceElementwise(
mask=p,
replacement=iap.Beta(0.5, 0.5) * 255,
per_channel=per_channel,
name=name,
deterministic=deterministic,
random_state=random_state
)
def CoarseSaltAndPepper(p=0, size_px=None, size_percent=None, per_channel=False, min_size=4, name=None,
deterministic=False, random_state=None):
"""
Adds coarse salt and pepper noise to an image, i.e. rectangles that contain noisy white-ish and black-ish pixels.
TODO replace dtype support with uint8 only, because replacement is geared towards that value range
dtype support::
See ``imgaug.augmenters.arithmetic.ReplaceElementwise``.
Parameters
----------
p : float or tuple of float or list of float or imgaug.parameters.StochasticParameter, optional
Probability of changing a pixel to salt/pepper noise.
* If a float, then that value will be used for all images as the
probability.
* If a tuple ``(a, b)``, then a probability will be sampled per image
from the range ``a <= x <= b``.
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then this parameter will be used as
the *mask*, i.e. it is expected to contain values between
0.0 and 1.0, where 1.0 means that salt/pepper is to be added
at that location.
size_px : int or tuple of int or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the noise
mask in absolute pixel dimensions.
* If an integer, then that size will be used for both height and
width. E.g. a value of 3 would lead to a ``3x3`` mask, which is then
upsampled to ``HxW``, where ``H`` is the image size and ``W`` the image width.
* If a tuple ``(a, b)``, then two values ``M``, ``N`` will be sampled from the
range ``[a..b]`` and the mask will be generated at size ``MxN``, then
upsampled to ``HxW``.
* If a StochasticParameter, then this parameter will be used to
determine the sizes. It is expected to be discrete.
size_percent : float or tuple of float or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the noise
mask *in percent* of the input image.
* If a float, then that value will be used as the percentage of the
height and width (relative to the original size). E.g. for value
p, the mask will be sampled from ``(p*H)x(p*W)`` and later upsampled
to ``HxW.``
* If a tuple ``(a, b)``, then two values ``m``, ``n`` will be sampled from the
interval ``(a, b)`` and used as the percentages, i.e the mask size
will be ``(m*H)x(n*W)``.
* If a StochasticParameter, then this parameter will be used to
sample the percentage values. It is expected to be continuous.
per_channel : bool or float, optional
Whether to use the same value (is dropped / is not dropped)
for all channels of a pixel (False) or to sample a new value for each
channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
min_size : int, optional
Minimum size of the low resolution mask, both width and height. If
`size_percent` or `size_px` leads to a lower value than this, `min_size`
will be used instead. This should never have a value of less than 2,
otherwise one may end up with a 1x1 low resolution mask, leading easily
to the whole image being replaced.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.CoarseSaltAndPepper(0.05, size_percent=(0.01, 0.1))
Replaces 5 percent of all pixels with salt/pepper in an image that has
1 to 10 percent of the input image size, then upscales the results
to the input image size, leading to large rectangular areas being replaced.
"""
mask = iap.handle_probability_param(p, "p", tuple_to_uniform=True, list_to_choice=True)
if size_px is not None:
mask_low = iap.FromLowerResolution(other_param=mask, size_px=size_px, min_size=min_size)
elif size_percent is not None:
mask_low = iap.FromLowerResolution(other_param=mask, size_percent=size_percent, min_size=min_size)
else:
raise Exception("Either size_px or size_percent must be set.")
replacement = iap.Beta(0.5, 0.5) * 255
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return ReplaceElementwise(
mask=mask_low,
replacement=replacement,
per_channel=per_channel,
name=name,
deterministic=deterministic,
random_state=random_state
)
def Salt(p=0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Adds salt noise to an image, i.e. white-ish pixels.
dtype support::
See ``imgaug.augmenters.arithmetic.ReplaceElementwise``.
Parameters
----------
p : float or tuple of float or list of float or imgaug.parameters.StochasticParameter, optional
Probability of changing a pixel to salt noise.
* If a float, then that value will be used for all images as the
probability.
* If a tuple ``(a, b)``, then a probability will be sampled per image
from the range ``a <= x <= b``.
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then this parameter will be used as
the *mask*, i.e. it is expected to contain values between
0.0 and 1.0, where 1.0 means that salt is to be added
at that location.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.Salt(0.05)
Replaces 5 percent of all pixels with salt.
"""
replacement01 = iap.ForceSign(
iap.Beta(0.5, 0.5) - 0.5,
positive=True,
mode="invert"
) + 0.5
replacement = replacement01 * 255 # FIXME max replacement seems to essentially never exceed 254
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return ReplaceElementwise(mask=p, replacement=replacement, per_channel=per_channel, name=name,
deterministic=deterministic, random_state=random_state)
def CoarseSalt(p=0, size_px=None, size_percent=None, per_channel=False, min_size=4, name=None, deterministic=False,
random_state=None):
"""
Adds coarse salt noise to an image, i.e. rectangles containing noisy white-ish pixels.
dtype support::
See ``imgaug.augmenters.arithmetic.ReplaceElementwise``.
Parameters
----------
p : float or tuple of float or list of float or imgaug.parameters.StochasticParameter, optional
Probability of changing a pixel to salt noise.
* If a float, then that value will be used for all images as the
probability.
* If a tuple ``(a, b)``, then a probability will be sampled per image
from the range ``a <= x <= b``.
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then this parameter will be used as
the *mask*, i.e. it is expected to contain values between
0.0 and 1.0, where 1.0 means that salt is to be added
at that location.
size_px : int or tuple of int or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the noise
mask in absolute pixel dimensions.
* If an integer, then that size will be used for both height and
width. E.g. a value of 3 would lead to a ``3x3`` mask, which is then
upsampled to ``HxW``, where H is the image size and W the image width.
* If a tuple ``(a, b)``, then two values M, N will be sampled from the
range ``[a..b]`` and the mask will be generated at size ``MxN``, then
upsampled to ``HxW``.
* If a StochasticParameter, then this parameter will be used to
determine the sizes. It is expected to be discrete.
size_percent : float or tuple of float or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the noise
mask *in percent* of the input image.
* If a float, then that value will be used as the percentage of the
height and width (relative to the original size). E.g. for value
p, the mask will be sampled from ``(p*H)x(p*W)`` and later upsampled
to ``HxW``.
* If a tuple ``(a, b)``, then two values ``m``, ``n`` will be sampled from the
interval ``(a, b)`` and used as the percentages, i.e the mask size
will be ``(m*H)x(n*W)``.
* If a StochasticParameter, then this parameter will be used to
sample the percentage values. It is expected to be continuous.
per_channel : bool or float, optional
Whether to use the same value (is dropped / is not dropped)
for all channels of a pixel (False) or to sample a new value for each
channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
min_size : int, optional
Minimum size of the low resolution mask, both width and height. If
`size_percent` or `size_px` leads to a lower value than this, `min_size`
will be used instead. This should never have a value of less than 2,
otherwise one may end up with a ``1x1`` low resolution mask, leading easily
to the whole image being replaced.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.CoarseSalt(0.05, size_percent=(0.01, 0.1))
Replaces 5 percent of all pixels with salt in an image that has
1 to 10 percent of the input image size, then upscales the results
to the input image size, leading to large rectangular areas being replaced.
"""
mask = iap.handle_probability_param(p, "p", tuple_to_uniform=True, list_to_choice=True)
if size_px is not None:
mask_low = iap.FromLowerResolution(other_param=mask, size_px=size_px, min_size=min_size)
elif size_percent is not None:
mask_low = iap.FromLowerResolution(other_param=mask, size_percent=size_percent, min_size=min_size)
else:
raise Exception("Either size_px or size_percent must be set.")
replacement01 = iap.ForceSign(
iap.Beta(0.5, 0.5) - 0.5,
positive=True,
mode="invert"
) + 0.5
replacement = replacement01 * 255
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return ReplaceElementwise(mask=mask_low, replacement=replacement, per_channel=per_channel, name=name,
deterministic=deterministic, random_state=random_state)
def Pepper(p=0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Adds pepper noise to an image, i.e. black-ish pixels.
This is similar to dropout, but slower and the black pixels are not uniformly black.
dtype support::
See ``imgaug.augmenters.arithmetic.ReplaceElementwise``.
Parameters
----------
p : float or tuple of float or list of float or imgaug.parameters.StochasticParameter, optional
Probability of changing a pixel to pepper noise.
* If a float, then that value will be used for all images as the
probability.
* If a tuple ``(a, b)``, then a probability will be sampled per image
from the range ``a <= x <= b``.
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then this parameter will be used as
the *mask*, i.e. it is expected to contain values between
0.0 and 1.0, where 1.0 means that pepper is to be added
at that location.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.Pepper(0.05)
Replaces 5 percent of all pixels with pepper.
"""
replacement01 = iap.ForceSign(
iap.Beta(0.5, 0.5) - 0.5,
positive=False,
mode="invert"
) + 0.5
replacement = replacement01 * 255
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return ReplaceElementwise(
mask=p,
replacement=replacement,
per_channel=per_channel,
name=name,
deterministic=deterministic,
random_state=random_state
)
def CoarsePepper(p=0, size_px=None, size_percent=None, per_channel=False, min_size=4, name=None, deterministic=False,
random_state=None):
"""
Adds coarse pepper noise to an image, i.e. rectangles that contain noisy black-ish pixels.
dtype support::
See ``imgaug.augmenters.arithmetic.ReplaceElementwise``.
Parameters
----------
p : float or tuple of float or list of float or imgaug.parameters.StochasticParameter, optional
Probability of changing a pixel to pepper noise.
* If a float, then that value will be used for all images as the
probability.
* If a tuple ``(a, b)``, then a probability will be sampled per image
from the range ``a <= x <= b.``
* If a list, then a random value will be sampled from that list
per image.
* If a StochasticParameter, then this parameter will be used as
the *mask*, i.e. it is expected to contain values between
0.0 and 1.0, where 1.0 means that pepper is to be added
at that location.
size_px : int or tuple of int or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the noise
mask in absolute pixel dimensions.
* If an integer, then that size will be used for both height and
width. E.g. a value of 3 would lead to a ``3x3`` mask, which is then
upsampled to ``HxW``, where ``H`` is the image size and W the image width.
* If a tuple ``(a, b)``, then two values ``M``, ``N`` will be sampled from the
range ``[a..b]`` and the mask will be generated at size ``MxN``, then
upsampled to ``HxW``.
* If a StochasticParameter, then this parameter will be used to
determine the sizes. It is expected to be discrete.
size_percent : float or tuple of float or imgaug.parameters.StochasticParameter, optional
The size of the lower resolution image from which to sample the noise
mask *in percent* of the input image.
* If a float, then that value will be used as the percentage of the
height and width (relative to the original size). E.g. for value
p, the mask will be sampled from ``(p*H)x(p*W)`` and later upsampled
to ``HxW``.
* If a tuple ``(a, b)``, then two values ``m``, ``n`` will be sampled from the
interval ``(a, b)`` and used as the percentages, i.e the mask size
will be ``(m*H)x(n*W)``.
* If a StochasticParameter, then this parameter will be used to
sample the percentage values. It is expected to be continuous.
per_channel : bool or float, optional
Whether to use the same value (is dropped / is not dropped)
for all channels of a pixel (False) or to sample a new value for each
channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
min_size : int, optional
Minimum size of the low resolution mask, both width and height. If
`size_percent` or `size_px` leads to a lower value than this, `min_size`
will be used instead. This should never have a value of less than 2,
otherwise one may end up with a 1x1 low resolution mask, leading easily
to the whole image being replaced.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.CoarsePepper(0.05, size_percent=(0.01, 0.1))
Replaces 5 percent of all pixels with pepper in an image that has
1 to 10 percent of the input image size, then upscales the results
to the input image size, leading to large rectangular areas being replaced.
"""
mask = iap.handle_probability_param(p, "p", tuple_to_uniform=True, list_to_choice=True)
if size_px is not None:
mask_low = iap.FromLowerResolution(other_param=mask, size_px=size_px, min_size=min_size)
elif size_percent is not None:
mask_low = iap.FromLowerResolution(other_param=mask, size_percent=size_percent, min_size=min_size)
else:
raise Exception("Either size_px or size_percent must be set.")
replacement01 = iap.ForceSign(
iap.Beta(0.5, 0.5) - 0.5,
positive=False,
mode="invert"
) + 0.5
replacement = replacement01 * 255
if name is None:
name = "Unnamed%s" % (ia.caller_name(),)
return ReplaceElementwise(
mask=mask_low,
replacement=replacement,
per_channel=per_channel,
name=name,
deterministic=deterministic,
random_state=random_state
)
class Invert(meta.Augmenter):
"""
Augmenter that inverts all values in images.
For the standard value range of 0-255 it converts 0 to 255, 255 to 0
and 10 to ``(255-10)=245``.
Let ``M`` be the maximum value possible, ``m`` the minimum value possible,
``v`` a value. Then the distance of ``v`` to ``m`` is ``d=abs(v-m)`` and the new value
is given by ``v'=M-d``.
dtype support::
if (min_value=None and max_value=None)::
* ``uint8``: yes; fully tested
* ``uint16``: yes; tested
* ``uint32``: yes; tested
* ``uint64``: yes; tested
* ``int8``: yes; tested
* ``int16``: yes; tested
* ``int32``: yes; tested
* ``int64``: yes; tested
* ``float16``: yes; tested
* ``float32``: yes; tested
* ``float64``: yes; tested
* ``float128``: yes; tested
* ``bool``: yes; tested
if (min_value!=None or max_value!=None)::
* ``uint8``: yes; fully tested
* ``uint16``: yes; tested
* ``uint32``: yes; tested
* ``uint64``: yes; tested
* ``int8``: yes; tested
* ``int16``: yes; tested
* ``int32``: yes; tested
* ``int64``: no (1)
* ``float16``: yes; tested
* ``float32``: yes; tested
* ``float64``: no (1)
* ``float128``: no (2)
* ``bool``: no (3)
- (1) Not allowed as int/float have to be increased in resolution when using min/max values.
- (2) Not tested.
- (3) Makes no sense when using min/max values.
Parameters
----------
p : float or imgaug.parameters.StochasticParameter, optional
The probability of an image to be inverted.
* If a float, then that probability will be used for all images.
* If a StochasticParameter, then that parameter will queried per
image and is expected to return values in the range ``[0.0, 1.0]``,
where values ``>0.5`` mean that the image/channel is supposed to be
inverted. Recommended to be some form of ``imgaug.parameters.Binomial``.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
min_value : None or number, optional
Minimum of the value range of input images, e.g. 0 for uint8 images.
If set to None, the value will be automatically derived from the image's dtype.
max_value : int or float, optional
Maximum of the value range of input images, e.g. 255 for uint8 images.
If set to None, the value will be automatically derived from the image's dtype.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.Invert(0.1)
Inverts the colors in 10 percent of all images.
>>> aug = iaa.Invert(0.1, per_channel=0.5)
For 50 percent of all images, it inverts all channels with a probability of
10 percent (same as the first example). For the other 50 percent of all
images, it inverts each channel individually with a probability of 10
percent (so some channels of an image may end up inverted, others not).
"""
# when no custom min/max are chosen, all bool, uint, int and float dtypes should be invertable (float tested only
# up to 64bit)
# when chosing custom min/max:
# - bool makes no sense, not allowed
# - int and float must be increased in resolution if custom min/max values are chosen,
# hence they are limited to 32 bit and below
# - float16 seems to not be perfectly accurate, but still ok-ish -- was off by 10 for center value of
# range (float 16 min, 16), where float 16 min is around -65500
ALLOW_DTYPES_CUSTOM_MINMAX = [
np.dtype(dt) for dt in [
np.uint8, np.uint16, np.uint32, np.uint64,
np.int8, np.int16, np.int32,
np.float16, np.float32
]
]
def __init__(self, p=0, per_channel=False, min_value=None, max_value=None, name=None, deterministic=False,
random_state=None):
super(Invert, self).__init__(name=name, deterministic=deterministic, random_state=random_state)
# TODO allow list and tuple for p
self.p = iap.handle_probability_param(p, "p")
self.per_channel = iap.handle_probability_param(per_channel, "per_channel")
self.min_value = min_value
self.max_value = max_value
self.dtype_kind_to_invert_func = {
"b": self._invert_bool,
"u": self._invert_uint,
"i": self._invert_int,
"f": self._invert_float
}
def _augment_images(self, images, random_state, parents, hooks):
nb_images = len(images)
nb_channels = meta.estimate_max_number_of_channels(images)
rss = ia.derive_random_states(random_state, 2)
p_samples = self.p.draw_samples((nb_images, nb_channels), random_state=rss[0])
per_channel_samples = self.per_channel.draw_samples((nb_images,), random_state=rss[1])
for image, per_channel_samples_i, p_samples_i in zip(images, per_channel_samples, p_samples):
min_value_dt, _, max_value_dt = iadt.get_value_range_of_dtype(image.dtype)
min_value = min_value_dt if self.min_value is None else self.min_value
max_value = max_value_dt if self.max_value is None else self.max_value
assert min_value >= min_value_dt,\
"Expected min_value to be above or equal to dtype's min value, got %s (vs. min possible %s for %s)" % (
str(min_value), str(min_value_dt), image.dtype.name)
assert max_value <= max_value_dt,\
"Expected max_value to be below or equal to dtype's max value, got %s (vs. max possible %s for %s)" % (
str(max_value), str(max_value_dt), image.dtype.name)
assert min_value < max_value, "Expected min_value to be below max_value, got %s and %s" % (
str(min_value), str(max_value))
if min_value != min_value_dt or max_value != max_value_dt:
ia.do_assert(image.dtype.type in self.ALLOW_DTYPES_CUSTOM_MINMAX,
"Can use custom min/max values only with the following dtypes: %s. Got: %s." % (
", ".join([dt.name for dt in self.ALLOW_DTYPES_CUSTOM_MINMAX]), image.dtype.name))
_invertfunc = self.dtype_kind_to_invert_func[image.dtype.kind]
if per_channel_samples_i > 0.5:
for c, p_sample in enumerate(p_samples_i):
if p_sample > 0.5:
image[..., c] = _invertfunc(image[..., c], min_value, max_value)
else:
if p_samples_i[0] > 0.5:
image[:, :, :] = _invertfunc(image, min_value, max_value)
return images
@classmethod
def _invert_bool(cls, arr, min_value, max_value):
ia.do_assert(min_value == 0, "Cannot modify min/max value for bool arrays in Invert.")
ia.do_assert(max_value == 1, "Cannot modify min/max value for bool arrays in Invert.")
return ~arr
@classmethod
def _invert_uint(cls, arr, min_value, max_value):
if min_value == 0 and max_value == np.iinfo(arr.dtype).max:
return max_value - arr
else:
return cls._invert_by_distance(
np.clip(arr, min_value, max_value),
min_value, max_value
)
@classmethod
def _invert_int(cls, arr, min_value, max_value):
# note that for int dtypes the max value is
# (-1) * min_value - 1
# e.g. -128 and 127 (min/max) for int8
# mapping example:
# [-4, -3, -2, -1, 0, 1, 2, 3]
# will be mapped to
# [ 3, 2, 1, 0, -1, -2, -3, -4]
# hence we can not simply compute the inverse as:
# after = (-1) * before
# but instead need
# after = (-1) * before - 1
# however, this exceeds the value range for the minimum value, e.g. for int8: -128 -> 128 -> 127,
# where 128 exceeds it. Hence, we must compute the inverse via a mask (extra step for the minimum)
# or we have to increase the resolution of the array. Here, a two-step approach is used.
if min_value == (-1) * max_value - 1:
mask = (arr == min_value)
# there is probably a one-liner here to do this, but
# ((-1) * (arr * ~mask) - 1) + mask * max_value
# has the disadvantage of inverting min_value to max_value - 1
# while
# ((-1) * (arr * ~mask) - 1) + mask * (max_value+1)
# ((-1) * (arr * ~mask) - 1) + mask * max_value + mask
# both sometimes increase the dtype resolution (e.g. int32 to int64)
n_min = np.sum(mask)
if n_min > 0:
arr[mask] = max_value
if n_min < arr.size:
arr[~mask] = (-1) * arr[~mask] - 1
return arr
else:
return cls._invert_by_distance(
np.clip(arr, min_value, max_value),
min_value, max_value
)
@classmethod
def _invert_float(cls, arr, min_value, max_value):
if np.isclose(max_value, (-1)*min_value, rtol=0):
return (-1) * arr
else:
return cls._invert_by_distance(
np.clip(arr, min_value, max_value),
min_value, max_value
)
@classmethod
def _invert_by_distance(cls, arr, min_value, max_value):
arr_modify = arr
if arr.dtype.kind in ["i", "f"]:
arr_modify = iadt.increase_array_resolutions_([np.copy(arr)], 2)[0]
distance_from_min = np.abs(arr_modify - min_value) # d=abs(v-min)
arr_modify = max_value - distance_from_min # v'=MAX-d
# due to floating point inaccuracies, we might exceed the min/max values for floats here, hence clip
# this happens especially for values close to the float dtype's maxima
if arr.dtype.kind == "f":
arr_modify = np.clip(arr_modify, min_value, max_value)
elif arr.dtype.kind in ["i", "f"]:
arr_modify = iadt.restore_dtypes_(arr_modify, [arr.dtype], clip=False)
return arr_modify
def _augment_heatmaps(self, heatmaps, random_state, parents, hooks):
return heatmaps
def _augment_keypoints(self, keypoints_on_images, random_state, parents, hooks):
return keypoints_on_images
def get_parameters(self):
return [self.p, self.per_channel, self.min_value, self.max_value]
# TODO remove from examples and mark as deprecated
def ContrastNormalization(alpha=1.0, per_channel=False, name=None, deterministic=False, random_state=None):
"""
Augmenter that changes the contrast of images.
dtype support:
See ``imgaug.augmenters.contrast.LinearContrast``.
Parameters
----------
alpha : number or tuple of number or list of number or imgaug.parameters.StochasticParameter, optional
Strength of the contrast normalization. Higher values than 1.0
lead to higher contrast, lower values decrease the contrast.
* If a number, then that value will be used for all images.
* If a tuple ``(a, b)``, then a value will be sampled per image from
the range ``a <= x <= b`` and be used as the alpha value.
* If a list, then a random value will be sampled per image from
that list.
* If a StochasticParameter, then this parameter will be used to
sample the alpha value per image.
per_channel : bool or float, optional
Whether to use the same value for all channels (False)
or to sample a new value for each channel (True).
If this value is a float ``p``, then for ``p`` percent of all images
`per_channel` will be treated as True, otherwise as False.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> iaa.ContrastNormalization((0.5, 1.5))
Decreases oder improves contrast per image by a random factor between
0.5 and 1.5. The factor 0.5 means that any difference from the center value
(i.e. 128) will be halved, leading to less contrast.
>>> iaa.ContrastNormalization((0.5, 1.5), per_channel=0.5)
Same as before, but for 50 percent of all images the normalization is done
independently per channel (i.e. factors can vary per channel for the same
image). In the other 50 percent of all images, the factor is the same for
all channels.
"""
# placed here to avoid cyclic dependency
from . import contrast as contrast_lib
return contrast_lib.LinearContrast(alpha=alpha, per_channel=per_channel, name=name, deterministic=deterministic,
random_state=random_state)
class JpegCompression(meta.Augmenter):
"""
Degrade image quality by applying JPEG compression to it.
During JPEG compression, high frequency components (e.g. edges) are removed.
With low compression (strength) only the highest frequency components are
removed, while very high compression (strength) will lead to only the lowest
frequency components "surviving". This lowers the image quality. For more
details, see https://en.wikipedia.org/wiki/Compression_artifact.
Note that this augmenter still returns images as numpy arrays (i.e. saves
the images with JPEG compression and then reloads them into arrays). It
does not return the raw JPEG file content.
dtype support::
* ``uint8``: yes; fully tested
* ``uint16``: ?
* ``uint32``: ?
* ``uint64``: ?
* ``int8``: ?
* ``int16``: ?
* ``int32``: ?
* ``int64``: ?
* ``float16``: ?
* ``float32``: ?
* ``float64``: ?
* ``float128``: ?
* ``bool``: ?
Parameters
----------
compression : number or tuple of number or list of number or \
imgaug.parameters.StochasticParameter, optional
Degree of compression used during jpeg compression within value range
``[0, 100]``. Higher values denote stronger compression and will cause
low-frequency components to disappear. Note that JPEG's compression
strength is also often set as a *quality*, which is the inverse of this
parameter. Common choices for the *quality* setting are around 80 to 95,
depending on the image. This translates here to a *compression*
parameter of around 20 to 5.
* If a single number, then that value will be used for the
compression degree.
* If a tuple of two number ``(a, b)``, then the compression will be
a value sampled from the interval ``[a..b]``.
* If a list, then a random value will be sampled and used as the
compression per image.
* If a StochasticParameter, then ``N`` samples will be drawn from
that parameter per ``N`` input images, each representing the
compression for the nth image. Expected to be discrete.
name : None or str, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
deterministic : bool, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
random_state : None or int or numpy.random.RandomState, optional
See :func:`imgaug.augmenters.meta.Augmenter.__init__`.
Examples
--------
>>> aug = iaa.JpegCompression(compression=(80, 95))
Removes high frequency components in images based on JPEG compression with
a *compression strength* between 80 and 95 (randomly sampled per image).
This corresponds to a (very low) *quality* setting of 5 to 20.
"""
def __init__(self, compression=50, name=None, deterministic=False, random_state=None):
super(JpegCompression, self).__init__(name=name, deterministic=deterministic, random_state=random_state)
# will be converted to int during augmentation, which is why we allow floats here
self.compression = iap.handle_continuous_param(compression, "compression", value_range=(0, 100),
tuple_to_uniform=True, list_to_choice=True)
# The value range 1 to 95 is suggested by PIL's save() documentation
# Values above 95 seem to not make sense (no improvement in visual quality, but large file size)
# A value of 100 would mostly deactivate jpeg compression
# A value of 0 would lead to no compression (instead of maximum compression)
# We use range 1 to 100 here, because this augmenter is about generating images for training
# and not for saving, hence we do not care about large file sizes
self.maximum_quality = 100
self.minimum_quality = 1
def _augment_images(self, images, random_state, parents, hooks):
result = images
nb_images = len(images)
samples = self.compression.draw_samples((nb_images,), random_state=random_state)
for i, (image, sample) in enumerate(zip(images, samples)):
ia.do_assert(image.dtype.name == "uint8", "Can apply jpeg compression only to uint8 images.")
nb_channels = image.shape[-1]
is_single_channel = (nb_channels == 1)
if is_single_channel:
image = image[..., 0]
sample = int(sample)
ia.do_assert(100 >= sample >= 0)
image_pil = PIL_Image.fromarray(image)
with tempfile.NamedTemporaryFile(mode="wb", suffix=".jpg") as f:
# Map from compression to quality used by PIL
# We have valid compressions from 0 to 100, i.e. 101 possible values
quality = int(
np.clip(
np.round(
self.minimum_quality
+ (self.maximum_quality - self.minimum_quality) * (1.0 - (sample / 101))
),
self.minimum_quality,
self.maximum_quality
)
)
image_pil.save(f, quality=quality)
if nb_channels == 1:
image = imageio.imread(f.name, pilmode="L")
else:
image = imageio.imread(f.name, pilmode="RGB")
if is_single_channel:
image = image[..., np.newaxis]
result[i] = image
return result
def _augment_heatmaps(self, heatmaps, random_state, parents, hooks):
return heatmaps
def _augment_keypoints(self, keypoints_on_images, random_state, parents, hooks):
return keypoints_on_images
def get_parameters(self):
return [self.compression]