},
rotate=(-10, 10),
shear=(-5, 5),
order=[0, 1],
cval=(0, 255),
mode=ia.ALL)),
iaa.SomeOf(
(0, 5),
[
#sometimes(iaa.Superpixels(p_replace=(0, 1.0), n_segments=(20, 200))),
iaa.OneOf([
iaa.GaussianBlur((0, 1.0)),
iaa.AverageBlur(k=(3, 5)),
#iaa.MedianBlur(k=(3, 5)),
]),
iaa.Sharpen(alpha=(0, 1.0), lightness=(0.9, 1.1)),
iaa.Emboss(alpha=(0, 1.0), strength=(0, 2.0)),
iaa.SimplexNoiseAlpha(
iaa.OneOf([
iaa.EdgeDetect(alpha=(0.5, 1.0)),
iaa.DirectedEdgeDetect(alpha=(0.5, 1.0),
direction=(0.0, 1.0)),
])),
iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, 0.01 * 255), per_channel=0.5),
iaa.OneOf([
iaa.Dropout((0.01, 0.05), per_channel=0.5),
iaa.CoarseDropout((0.01, 0.03),
size_percent=(0.01, 0.02),
per_channel=0.2),
]),
]),
iaa.OneOf([
iaa.MedianBlur(k=(3, 7)),
iaa.BilateralBlur(
d=(3, 10), sigma_color=(10, 250), sigma_space=(10, 250)),
iaa.MotionBlur(k=(3, 9), angle=[-45, 45]),
iaa.MeanShiftBlur(spatial_radius=(5.0, 10.0),
color_radius=(5.0, 10.0)),
iaa.AllChannelsCLAHE(clip_limit=(1, 10)),
iaa.AllChannelsHistogramEqualization(),
iaa.GammaContrast((0.5, 1.5), per_channel=True),
iaa.GammaContrast((0.5, 1.5)),
iaa.SigmoidContrast(gain=(3, 10), cutoff=(0.4, 0.6), per_channel=True),
iaa.SigmoidContrast(gain=(3, 10), cutoff=(0.4, 0.6)),
iaa.HistogramEqualization(),
iaa.Sharpen(alpha=0.5)
]),
iaa.OneOf([
iaa.AveragePooling([2, 3]),
iaa.MaxPooling(([2, 3], [2, 3])),
]),
iaa.OneOf([
iaa.Clouds(),
iaa.Snowflakes(flake_size=(0.1, 0.4), speed=(0.01, 0.05)),
iaa.Rain(speed=(0.1, 0.3))
])
],
random_order=True)
def get_color_augmentation(augment_prob):
config = process_config("./config.json")
args = Bunch(config.pretrained_model)
model = face_model.FaceModel(args)
sometimes = lambda aug: iaa.Sometimes(0.8, aug)
seq = iaa.Sequential([
iaa.Fliplr(0.5),
sometimes(
iaa.OneOf([
iaa.Grayscale(alpha=(0.0, 1.0)),
iaa.AddToHueAndSaturation((-20, 20)),
iaa.Add((-20, 20), per_channel=0.5),
iaa.Multiply((0.5, 1.5), per_channel=0.5),
iaa.GaussianBlur((0, 2.0)),
iaa.ContrastNormalization((0.5, 2.0), per_channel=0.5),
iaa.Sharpen(alpha=(0, 0.5), lightness=(0.7, 1.3)),
iaa.Emboss(alpha=(0, 0.5), strength=(0, 1.5))
]))
])
for mset in ['train', 'test', 'testcam']:
output_dir = './data/embedding/%s/%s' % (
args.model.split(',')[0].split('/')[-2], mset)
if not os.path.exists(output_dir):
os.makedirs(output_dir)
for rdir, sdir, files in os.walk('./datasets/aligned/%s/112x112/' % mset):
for file in tqdm(files):
if file == '.' or file == '..':
continue
#fn, fe = os.path.splitext(file)
def _apply_aug_default():
"""
Apply General augmentation in a single pipeline
"""
sometimes = (lambda aug: iaa.Sometimes(0.5, aug))
# Define our sequence of augmentation steps that will be applied to every image.
return iaa.Sequential(
[
#
# Apply the following augmenters to most images.
#
iaa.Fliplr(0.5), # horizontally flip 50% of all images
iaa.Flipud(0.2), # vertically flip 20% of all images
# crop some of the images by 0-10% of their height/width
sometimes(iaa.Crop(percent=(0, 0.1))),
# Apply affine transformations to some of the images
# - scale to 80-120% of image height/width (each axis independently)
# - translate by -20 to +20 relative to height/width (per axis)
# - rotate by -45 to +45 degrees
# - shear by -16 to +16 degrees
# - order: use nearest neighbour or bilinear interpolation (fast)
# - mode: use any available mode to fill newly created pixels
# see API or scikit-image for which modes are available
# - cval: if the mode is constant, then use a random brightness
# for the newly created pixels (e.g. sometimes black,
# sometimes white)
sometimes(
iaa.Affine(scale={
"x": (0.8, 1.2),
"y": (0.8, 1.2)
},
translate_percent={
"x": (-0.2, 0.2),
"y": (-0.2, 0.2)
},
rotate=(-45, 45),
shear=(-16, 16),
order=[0, 1],
cval=(0, 255),
mode=ia.ALL)),
#
# Execute 0 to 5 of the following (less important) augmenters per
# image. Don't execute all of them, as that would often be way too
# strong.
#
iaa.SomeOf(
(0, 5),
[
# Convert some images into their superpixel representation,
# sample between 20 and 200 superpixels per image, but do
# not replace all superpixels with their average, only
# some of them (p_replace).
sometimes(
iaa.Superpixels(p_replace=(0, 1.0),
n_segments=(20, 200))),
# Blur each image with varying strength using
# gaussian blur (sigma between 0 and 3.0),
# average/uniform blur (kernel size between 2x2 and 7x7)
# median blur (kernel size between 3x3 and 11x11).
iaa.OneOf([
iaa.GaussianBlur((0, 3.0)),
iaa.AverageBlur(k=(2, 7)),
iaa.MedianBlur(k=(3, 11)),
]),
# Sharpen each image, overlay the result with the original
# image using an alpha between 0 (no sharpening) and 1
# (full sharpening effect).
iaa.Sharpen(alpha=(0, 1.0), lightness=(0.75, 1.5)),
# Same as sharpen, but for an embossing effect.
iaa.Emboss(alpha=(0, 1.0), strength=(0, 2.0)),
# Search in some images either for all edges or for
# directed edges. These edges are then marked in a black
# and white image and overlayed with the original image
# using an alpha of 0 to 0.7.
sometimes(
iaa.OneOf([
iaa.EdgeDetect(alpha=(0, 0.7)),
iaa.DirectedEdgeDetect(alpha=(0, 0.7),
direction=(0.0, 1.0)),
])),
# Add gaussian noise to some images.
# In 50% of these cases, the noise is randomly sampled per
# channel and pixel.
# In the other 50% of all cases it is sampled once per
# pixel (i.e. brightness change).
iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, 0.05 * 255), per_channel=0.5),
# Either drop randomly 1 to 10% of all pixels (i.e. set
# them to black) or drop them on an image with 2-5% percent
# of the original size, leading to large dropped
# rectangles.
iaa.OneOf([
iaa.Dropout((0.01, 0.1), per_channel=0.5),
iaa.CoarseDropout((0.03, 0.15),
size_percent=(0.02, 0.05),
per_channel=0.2),
]),
# Invert each image's channel with 5% probability.
# This sets each pixel value v to 255-v.
iaa.Invert(0.05,
per_channel=True), # invert color channels
# Add a value of -10 to 10 to each pixel.
iaa.Add((-10, 10), per_channel=0.5),
# Change brightness of images (50-150% of original value).
iaa.Multiply((0.5, 1.5), per_channel=0.5),
# Improve or worsen the contrast of images.
iaa.LinearContrast((0.5, 2.0), per_channel=0.5),
# Convert each image to grayscale and then overlay the
# result with the original with random alpha. I.e. remove
# colors with varying strengths.
iaa.Grayscale(alpha=(0.0, 1.0)),
# In some images move pixels locally around (with random
# strengths).
sometimes(
iaa.ElasticTransformation(alpha=(0.5, 3.5), sigma=0.25)
),
# In some images distort local areas with varying strength.
sometimes(iaa.PiecewiseAffine(scale=(0.01, 0.05)))
],
# do all of the above augmentations in random order
random_order=True)
],
# do all of the above augmentations in random order
random_order=True)
import sys
import random
import numpy as np
import cv2
import skimage.io
import warnings
warnings.simplefilter('ignore')
import time
####################################################################
# Augmentations
####################################################################
# TGAR, add your own augmentations using the iaa package
# Sharpen and Emboss
SEaug = iaa.Sequential([
iaa.Sharpen(alpha=(0.0, 1.0), lightness=(0.75, 2.0)),
iaa.Emboss(alpha=(0.0, 1.0), strength=(0.5, 1.5))
],
random_order=True)
# Gaussian Noise
GNaug = iaa.AdditiveGaussianNoise(scale=(0, 0.05 * 255))
# Color
Caug = iaa.Sequential([
iaa.WithChannels(0, iaa.Add((10, 100))),
iaa.WithChannels(1, iaa.Add((10, 100))),
iaa.WithChannels(2, iaa.Add((10, 100)))
])
# Brightness and Contrast
def test_pickleable(self):
aug = iaa.Sharpen(alpha=(0.0, 1.0), lightness=(1, 3), seed=1)
runtest_pickleable_uint8_img(aug, iterations=20)
def __init__(self,aug_level):
frequently = lambda aug: iaa.Sometimes(0.9, aug)
often = lambda aug: iaa.Sometimes(0.7, aug)
sometimes = lambda aug: iaa.Sometimes(0.5, aug)
occasionally = lambda aug: iaa.Sometimes(0.3, aug)
rarely = lambda aug: iaa.Sometimes(0.1, aug)
if aug_level == 'high':
self.aug = iaa.Sequential([
sometimes(iaa.Crop(percent=(0, 0.1))),
rarely(iaa.ContrastNormalization((0.75, 1.5))),
occasionally(iaa.SigmoidContrast(gain=(3, 10), cutoff=(0.1, 0.2))),
sometimes(iaa.Affine(
scale={"x": (0.8, 1.2), "y": (0.8, 1.2)},
translate_percent={"x": (-0.2, 0.2), "y": (-0.2, 0.2)},
rotate=(-20, 20),
shear=(-8, 8),
order=[0, 1],
cval=(0, 255),
mode=ia.ALL
)),
iaa.SomeOf((0, 5),
[
# Blur each image with varying strength using
# gaussian blur (sigma between 0 and 3.0),
# average/uniform blur (kernel size between 2x2 and 7x7)
# median blur (kernel size between 3x3 and 11x11).
iaa.OneOf([
iaa.GaussianBlur((0, 3.0)),
iaa.AverageBlur(k=(2, 7)),
iaa.MedianBlur(k=(3, 11)),
]),
# Sharpen each image, overlay the result with the original
# image using an alpha between 0 (no sharpening) and 1
# (full sharpening effect).
iaa.Sharpen(alpha=(0, 1.0), lightness=(0.75, 1.5)),
# Search in some images either for all edges or for
# directed edges. These edges are then marked in a black
# and white image and overlayed with the original image
# using an alpha of 0 to 0.7.
sometimes(iaa.OneOf([
iaa.EdgeDetect(alpha=(0, 0.7)),
iaa.DirectedEdgeDetect(
alpha=(0, 0.7), direction=(0.0, 1.0)
),
])),
# Add gaussian noise to some images.
# In 50% of these cases, the noise is randomly sampled per
# channel and pixel.
# In the other 50% of all cases it is sampled once per
# pixel (i.e. brightness change).
iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, 0.03*255), per_channel=0.5
),
# Either drop randomly 1 to 10% of all pixels (i.e. set
# them to black) or drop them on an image with 2-5% percent
# of the original size, leading to large dropped
# rectangles.
iaa.OneOf([
iaa.Dropout((0.01, 0.1), per_channel=0.5),
iaa.CoarseDropout(
(0.03, 0.15), size_percent=(0.01, 0.02),
per_channel=0.2
),
]),
# Invert each image's channel with 5% probability.
# This sets each pixel value v to 255-v.
iaa.Invert(0.05, per_channel=True), # invert color channels
# Add a value of -10 to 10 to each pixel.
iaa.Add((-10, 10), per_channel=0.5),
# Change brightness of images (50-150% of original value).
iaa.Multiply((0.5, 1.0), per_channel=0.5),
# Improve or worsen the contrast of images.
iaa.LinearContrast((0.5, 2.0), per_channel=0.5),
# In some images move pixels locally around (with random
# strengths).
sometimes(
iaa.ElasticTransformation(alpha=(0.5, 3.5), sigma=0.25)
),
# In some images distort local areas with varying strength.
sometimes(iaa.PiecewiseAffine(scale=(0.01, 0.05)))
], random_order=True)
], random_order=True)
elif aug_level == 'low':
self.aug = iaa.Sequential([
iaa.OneOf([
iaa.Sometimes(0.5, iaa.AdditiveGaussianNoise(scale=(0.0, 0.1*255))),
iaa.Sometimes(0.1, iaa.AdditiveGaussianNoise(scale=(0.0, 0.3*255)))
]),
iaa.Sometimes(0.5, iaa.GammaContrast((0.5, 2.0))),
iaa.Sometimes(0.5, iaa.imgcorruptlike.Brightness(severity=2)),
iaa.Sometimes(0.5, iaa.Affine(
rotate=(-20, 20),
)),
], random_order=True)
else:
raise NotImplementedError
def sharpen(img):
seq = iaa.Sequential([iaa.Sharpen(alpha=(0, 1.0), lightness=(0.75, 1.5))])
return seq.augment_image(img)
from imgaug import augmenters as iaa
from keras.regularizers import l2
#Data augmentation
aug90 = iaa.Sometimes(
1, iaa.Affine(rotate=(-90, -90), mode="constant", cval=(-1000, -1000)))
aug180 = iaa.Sometimes(
1, iaa.Affine(rotate=(-180, -180), mode="constant", cval=(-1000, -1000)))
aug270 = iaa.Sometimes(
1, iaa.Affine(rotate=(-270, -270), mode="constant", cval=(-1000, -1000)))
augFlip1 = iaa.Sometimes(1, iaa.Fliplr(1))
augFlip2 = iaa.Sometimes(1, iaa.Flipud(1))
augCon = iaa.Sometimes(1, iaa.ContrastNormalization(0.5, 0.9))
augBlur = iaa.Sometimes(1, iaa.GaussianBlur(sigma=(0.0, 0.8)))
augSharpen = iaa.Sometimes(1, iaa.Sharpen(alpha=0.1, lightness=0.7))
def augmentHK(image):
w1 = image.reshape(image.shape[0], image.shape[1], image.shape[2])
k = random.randrange(1, 11)
global w2
if k == 1:
w2 = aug90.augment_image(w1)
elif k == 2:
w2 = aug180.augment_image(w1)
elif k == 3:
w2 = aug270.augment_image(w1)
elif k == 4:
w2 = augFlip1.augment_image(w1)
elif k == 5:
class DetectorDataset(utils.Dataset):
"""Dataset class for training pneumonia detection on the RSNA pneumonia dataset.
"""
augmentation_channel_1 = iaa.Sequential([iaa.Sharpen(alpha=0.5, lightness=2.0)])
augmentation_channel_2 = iaa.Sequential([iaa.ContrastNormalization((2, 1))])
#MASK_AUGMENTERS = ["Sequential", "SomeOf", "OneOf", "Sometimes", "Fliplr", "Flipud", "CropAndPad", "Affine", "PiecewiseAffine"]
augmentation_mode = False
def __init__(self, image_fps, image_annotations, orig_height, orig_width):
super().__init__(self)
# Add classes
self.add_class('pneumonia', 1, 'Lung Opacity')
# add images
for i, fp in enumerate(image_fps):
annotations = image_annotations[fp]
self.add_image('pneumonia', image_id=i, path=fp,
annotations=annotations, orig_height=orig_height, orig_width=orig_width)
def image_reference(self, image_id):
info = self.image_info[image_id]
return info['path']
def load_image(self, image_id):
info = self.image_info[image_id]
fp = info['path']
ds = pydicom.read_file(fp)
image = ds.pixel_array
if len(image.shape) != 3 or image.shape[2] != 3:
if self.augmentation_mode:
#add augmentation as 2 of 3 channels
# Store shapes before augmentation to compare
image_shape = image.shape
channel_2 = self.augmentation_channel_1.augment_image(image)
channel_3 = self.augmentation_channel_2.augment_image(image)
image = np.stack((image, channel_2, channel_3), -1)
else:
image = np.stack((image,)*3, -1)
return image
def load_mask(self, image_id):
info = self.image_info[image_id]
annotations = info['annotations']
count = len(annotations)
if count == 0:
mask = np.zeros((info['orig_height'], info['orig_width'], 1), dtype=np.uint8)
class_ids = np.zeros((1,), dtype=np.int32)
else:
mask = np.zeros((info['orig_height'], info['orig_width'], count), dtype=np.uint8)
class_ids = np.zeros((count,), dtype=np.int32)
if self.augmentation_mode:
def hook(images, augmenter, parents, default):
"""Determines which augmenters to apply to masks."""
return augmenter.__class__.__name__ in self.MASK_AUGMENTERS
for i, a in enumerate(annotations):
if a['Target'] == 1:
x = int(a['x'])
y = int(a['y'])
w = int(a['width'])
h = int(a['height'])
mask_instance = mask[:, :, i].copy()
cv2.rectangle(mask_instance, (x, y), (x + w, y + h), 255, -1)
mask[:, :, i] = mask_instance
class_ids[i] = 1
return mask.astype(np.bool), class_ids.astype(np.int32)
FTR_CLR_HIST = True
COLORSPACE = 'YCrCb'
ORIENT = 9
PIXEL_PER_CELL = 8
CELL_PER_BLOCK = 2
HOG_CHNL = 'ALL'
SPATIAL_SIZE = (32, 32)
HISTO_BINS = 32
BINS_RANGE = (0, 256)
SEQ_RTT = iaa.Affine(rotate=(-20, 20)) # rotate by -20 to +20 degrees
SEQ_FLIP = iaa.Fliplr(1.0) # always horizontally flip each input image
SEQ_BLR = iaa.GaussianBlur(sigma=(0,
3.0)) # blur images with a sigma of 0 to 3.0
SEQ_SHRP = iaa.Sharpen(alpha=(0, 1.0), lightness=(0.75, 1.5)) # sharpen images
SEQ_BRT = iaa.Add(
(-10, 10), per_channel=0.5
) # change brightness of images (by -10 to 10 of original value)
VEHICLE_IMG_PATHS = 'C:\\Users\\Yiqun\Downloads\\training_images\\vehicles\\*\\*.png'
NON_VEHICLE_IMG_PATHS = 'C:\\Users\\Yiqun\Downloads\\training_images\\non-vehicles\\*\\*.png'
def extract_hog_features(img,
orient,
pix_per_cell,
cell_per_block,
vis=False,
feature_vec=True):
#if call with two outputs then vis = True
def __init__(self,
ann_file,
root,
remove_images_without_annotations,
mask_type='polygon',
transforms=None):
# "mask_type" = "polygon" or "image"
self.mask_type = mask_type
self.img_key_list = list()
self.img_dict = dict()
self.ann_info = dict()
cls = RSNADataset.CLASSES
self.class_to_ind = dict(zip(cls, range(len(cls))))
for dirName, subdirList, fileList in os.walk(root):
for filename in fileList:
filename, ext = os.path.splitext(filename)
if ext.lower() in [".png", ".jpg", ".jpeg"]:
self.img_dict[filename] = os.path.join(
dirName, filename + ext)
self.ann_info[filename] = list()
# csv 용도 이며, mask 이미지인 경우 다르게 작업
with open(ann_file, 'r') as ann_f:
ann_cvf = csv.reader(ann_f)
# patientId,x,y,width,height,Target
for i, line in enumerate(ann_cvf):
if i == 0:
continue
filename, x, y, w, h, target = line
target = int(target)
if remove_images_without_annotations:
if target == 0:
continue
x1 = int(x)
y1 = int(y)
w = int(w)
h = int(h)
x2 = x1 + w
y2 = y1 + h
self.img_key_list.append(filename)
else:
self.img_key_list.append(filename)
x1 = 0
y1 = 0
x2 = 0
y2 = 0
try:
self.ann_info[filename].append([x1, y1, x2, y2, target])
except KeyError:
continue
self.augmentation = iaa.Sequential([
iaa.OneOf([ ## geometric transform
iaa.Affine(
scale={
"x": (0.98, 1.02),
"y": (0.98, 1.02)
},
translate_percent={
"x": (-0.02, 0.02),
"y": (-0.04, 0.04)
},
rotate=(-2, 2),
shear=(-1, 1),
),
iaa.PiecewiseAffine(scale=(0.001, 0.025)),
]),
iaa.OneOf([ ## brightness or contrast
iaa.Multiply((0.9, 1.1)),
iaa.ContrastNormalization((0.9, 1.1)),
]),
iaa.OneOf([ ## blur or sharpen
iaa.GaussianBlur(sigma=(0.0, 0.1)),
iaa.Sharpen(alpha=(0.0, 0.1)),
]),
])
# 중복 방지 RSNA csv 파일 참조
self.img_key_list = list(set(self.img_key_list))
self.transforms = transforms
def transform(aug_type, magnitude, X):
if aug_type == "crop":
X_aug = iaa.Crop(px=(0, int(magnitude * 32))).augment_images(X)
elif aug_type == "gaussian-blur":
X_aug = iaa.GaussianBlur(sigma=(0, magnitude * 25.0)).augment_images(X)
elif aug_type == "rotate":
X_aug = iaa.Affine(rotate=(-180 * magnitude, 180 * magnitude)).augment_images(X)
elif aug_type == "shear":
X_aug = iaa.Affine(shear=(-90 * magnitude, 90 * magnitude)).augment_images(X)
elif aug_type == "translate-x":
X_aug = iaa.Affine(
translate_percent={"x": (-magnitude, magnitude), "y": (0, 0)}
).augment_images(X)
elif aug_type == "translate-y":
X_aug = iaa.Affine(
translate_percent={"x": (0, 0), "y": (-magnitude, magnitude)}
).augment_images(X)
elif aug_type == "horizontal-flip":
X_aug = iaa.Fliplr(magnitude).augment_images(X)
elif aug_type == "vertical-flip":
X_aug = iaa.Flipud(magnitude).augment_images(X)
elif aug_type == "sharpen":
X_aug = iaa.Sharpen(
alpha=(0, 1.0), lightness=(0.50, 5 * magnitude)
).augment_images(X)
elif aug_type == "emboss":
X_aug = iaa.Emboss(
alpha=(0, 1.0), strength=(0.0, 20.0 * magnitude)
).augment_images(X)
elif aug_type == "additive-gaussian-noise":
X_aug = iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, magnitude * 255), per_channel=0.5
).augment_images(X)
elif aug_type == "dropout":
X_aug = iaa.Dropout(
(0.01, max(0.011, magnitude)), per_channel=0.5
).augment_images(
X
) # Dropout first argument should be smaller than second one
elif aug_type == "coarse-dropout":
X_aug = iaa.CoarseDropout(
(0.03, 0.15), size_percent=(0.30, np.log10(magnitude * 3)), per_channel=0.2
).augment_images(X)
elif aug_type == "gamma-contrast":
X_norm = normalize(X)
X_aug_norm = iaa.GammaContrast(magnitude * 1.75).augment_images(
X_norm
) # needs 0-1 values
X_aug = denormalize(X_aug_norm)
elif aug_type == "brighten":
X_aug = iaa.Add(
(int(-40 * magnitude), int(40 * magnitude)), per_channel=0.5
).augment_images(
X
) # brighten
elif aug_type == "invert":
X_aug = iaa.Invert(1.0).augment_images(X) # magnitude not used
elif aug_type == "fog":
X_aug = iaa.Fog().augment_images(X) # magnitude not used
elif aug_type == "clouds":
X_aug = iaa.Clouds().augment_images(X) # magnitude not used
elif aug_type == "histogram-equalize":
X_aug = iaa.AllChannelsHistogramEqualization().augment_images(
X
) # magnitude not used
elif aug_type == "super-pixels": # deprecated
X_norm = normalize(X)
X_norm2 = (X_norm * 2) - 1
X_aug_norm2 = iaa.Superpixels(
p_replace=(0, magnitude), n_segments=(100, 100)
).augment_images(X_norm2)
X_aug_norm = (X_aug_norm2 + 1) / 2
X_aug = denormalize(X_aug_norm)
elif aug_type == "perspective-transform":
X_norm = normalize(X)
X_aug_norm = iaa.PerspectiveTransform(
scale=(0.01, max(0.02, magnitude))
).augment_images(
X_norm
) # first scale param must be larger
np.clip(X_aug_norm, 0.0, 1.0, out=X_aug_norm)
X_aug = denormalize(X_aug_norm)
elif aug_type == "elastic-transform": # deprecated
X_norm = normalize(X)
X_norm2 = (X_norm * 2) - 1
X_aug_norm2 = iaa.ElasticTransformation(
alpha=(0.0, max(0.5, magnitude * 300)), sigma=5.0
).augment_images(X_norm2)
X_aug_norm = (X_aug_norm2 + 1) / 2
X_aug = denormalize(X_aug_norm)
elif aug_type == "add-to-hue-and-saturation":
X_aug = iaa.AddToHueAndSaturation(
(int(-45 * magnitude), int(45 * magnitude))
).augment_images(X)
elif aug_type == "coarse-salt-pepper":
X_aug = iaa.CoarseSaltAndPepper(p=0.2, size_percent=magnitude).augment_images(X)
elif aug_type == "grayscale":
X_aug = iaa.Grayscale(alpha=(0.0, magnitude)).augment_images(X)
else:
raise ValueError
return X_aug
X_train = X_train.reshape(-1, 28, 28, 1)
#Single Training Image
img = images.iloc[0].as_matrix().reshape((28, 28))
plt.imshow(img, cmap='binary')
plt.savefig('real.png')
plt.show()
########################################################################################################
#Agumented Images
X_train_aug1 = iaa.Fliplr(1).augment_images(X_train)
X_train_aug2 = iaa.AverageBlur(k=(3, 4)).augment_images(X_train)
X_train_aug3 = iaa.Emboss(alpha=(0.8, 1.0),
strength=(0, 3.0)).augment_images(X_train)
X_train_aug4 = iaa.Sharpen(alpha=(0.8, 0.9),
lightness=(0.75, 1.5)).augment_images(X_train)
X_train_aug5 = iaa.AdditiveGaussianNoise(
loc=0, scale=(5.0, 0.05 * 255)).augment_images(X_train)
X_train_aug6 = iaa.ContrastNormalization((2.0, 2.5)).augment_images(X_train)
for i in range(10):
#visualize X_train[i], X_train_aug1[i], etc
gs = gridspec.GridSpec(3, 3)
ax = plt.subplot(gs[0, 0])
imgplot = ax.imshow(X_train[i].reshape((28, 28)), cmap="binary")
ax.set_xticks([])
ax.set_yticks([])
ax.set_title('Original')
ax.title.set_size(10)
def test_alpha_075(self):
aug = iaa.Sharpen(alpha=0.75, lightness=1)
observed = aug.augment_image(self.base_img)
expected = self._compute_sharpened_base_img(
0.75 * 1, 0.25 * self.m_noop + 0.75 * self.m)
assert np.allclose(observed, expected)
def draw_per_augmenter_images():
print("[draw_per_augmenter_images] Loading image...")
image = misc.imresize(
ndimage.imread("quokka.jpg")[0:643, 0:643], (128, 128))
#image = misc.imresize(data.chelsea()[0:300, 50:350, :], (128, 128))
#image = misc.imresize(data.astronaut(), (128, 128))
#keypoints = [ia.Keypoint(x=43, y=43), ia.Keypoint(x=78, y=40), ia.Keypoint(x=64, y=73)] # left eye, right eye, mouth
keypoints = [
ia.Keypoint(x=34, y=15),
ia.Keypoint(x=85, y=13),
ia.Keypoint(x=63, y=73)
] # left ear, right ear, mouth
keypoints = [ia.KeypointsOnImage(keypoints, shape=image.shape)]
print("[draw_per_augmenter_images] Initializing...")
rows_augmenters = [
("Noop", [("", iaa.Noop()) for _ in sm.xrange(5)]),
#("Crop", [iaa.Crop(px=vals) for vals in [(2, 4), (4, 8), (6, 16), (8, 32), (10, 64)]]),
("Crop\n(top, right,\nbottom, left)", [
(str(vals), iaa.Crop(px=vals))
for vals in [(2, 0, 0, 0), (0, 8, 8, 0), (4, 0, 16,
4), (8, 0, 0,
32), (32, 64, 0, 0)]
]),
("Fliplr", [(str(p), iaa.Fliplr(p)) for p in [0, 0, 1, 1, 1]]),
("Flipud", [(str(p), iaa.Flipud(p)) for p in [0, 0, 1, 1, 1]]),
("Superpixels\np_replace=1", [("n_segments=%d" % (n_segments, ),
iaa.Superpixels(p_replace=1.0,
n_segments=n_segments))
for n_segments in [25, 50, 75, 100, 125]
]),
("Superpixels\nn_segments=100",
[("p_replace=%.2f" % (p_replace, ),
iaa.Superpixels(p_replace=p_replace, n_segments=100))
for p_replace in [0, 0.25, 0.5, 0.75, 1.0]]),
("Invert", [("p=%d" % (p, ), iaa.Invert(p=p))
for p in [0, 0, 1, 1, 1]]),
("Invert\n(per_channel)", [("p=%.2f" % (p, ),
iaa.Invert(p=p, per_channel=True))
for p in [0.5, 0.5, 0.5, 0.5, 0.5]]),
("Add", [("value=%d" % (val, ), iaa.Add(val))
for val in [-45, -25, 0, 25, 45]]),
("Add\n(per channel)", [
("value=(%d, %d)" % (
vals[0],
vals[1],
), iaa.Add(vals, per_channel=True))
for vals in [(-55, -35), (-35, -15), (-10, 10), (15, 35), (35, 55)]
]),
("Multiply", [("value=%.2f" % (val, ), iaa.Multiply(val))
for val in [0.25, 0.5, 1.0, 1.25, 1.5]]),
("Multiply\n(per channel)",
[("value=(%.2f, %.2f)" % (
vals[0],
vals[1],
), iaa.Multiply(vals, per_channel=True))
for vals in [(0.15, 0.35), (0.4, 0.6), (0.9, 1.1), (1.15,
1.35), (1.4,
1.6)]]),
("GaussianBlur", [("sigma=%.2f" % (sigma, ),
iaa.GaussianBlur(sigma=sigma))
for sigma in [0.25, 0.50, 1.0, 2.0, 4.0]]),
("AverageBlur", [("k=%d" % (k, ), iaa.AverageBlur(k=k))
for k in [1, 3, 5, 7, 9]]),
("MedianBlur", [("k=%d" % (k, ), iaa.MedianBlur(k=k))
for k in [1, 3, 5, 7, 9]]),
("Sharpen\n(alpha=1)", [("lightness=%.2f" % (lightness, ),
iaa.Sharpen(alpha=1, lightness=lightness))
for lightness in [0, 0.5, 1.0, 1.5, 2.0]]),
("Emboss\n(alpha=1)", [("strength=%.2f" % (strength, ),
iaa.Emboss(alpha=1, strength=strength))
for strength in [0, 0.5, 1.0, 1.5, 2.0]]),
("EdgeDetect", [("alpha=%.2f" % (alpha, ), iaa.EdgeDetect(alpha=alpha))
for alpha in [0.0, 0.25, 0.5, 0.75, 1.0]]),
("DirectedEdgeDetect\n(alpha=1)",
[("direction=%.2f" % (direction, ),
iaa.DirectedEdgeDetect(alpha=1, direction=direction))
for direction in [
0.0, 1 * (360 / 5) / 360, 2 * (360 / 5) / 360, 3 * (360 / 5) /
360, 4 * (360 / 5) / 360
]]),
("AdditiveGaussianNoise",
[("scale=%.2f*255" % (scale, ),
iaa.AdditiveGaussianNoise(scale=scale * 255))
for scale in [0.025, 0.05, 0.1, 0.2, 0.3]]),
("AdditiveGaussianNoise\n(per channel)",
[("scale=%.2f*255" % (scale, ),
iaa.AdditiveGaussianNoise(scale=scale * 255, per_channel=True))
for scale in [0.025, 0.05, 0.1, 0.2, 0.3]]),
("Dropout", [("p=%.2f" % (p, ), iaa.Dropout(p=p))
for p in [0.025, 0.05, 0.1, 0.2, 0.4]]),
("Dropout\n(per channel)", [("p=%.2f" % (p, ),
iaa.Dropout(p=p, per_channel=True))
for p in [0.025, 0.05, 0.1, 0.2, 0.4]]),
("CoarseDropout\n(p=0.2)",
[("size_percent=%.2f" % (size_percent, ),
iaa.CoarseDropout(p=0.2, size_percent=size_percent, min_size=2))
for size_percent in [0.3, 0.2, 0.1, 0.05, 0.02]]),
("CoarseDropout\n(p=0.2, per channel)",
[("size_percent=%.2f" % (size_percent, ),
iaa.CoarseDropout(p=0.2,
size_percent=size_percent,
per_channel=True,
min_size=2))
for size_percent in [0.3, 0.2, 0.1, 0.05, 0.02]]),
("ContrastNormalization", [("alpha=%.1f" % (alpha, ),
iaa.ContrastNormalization(alpha=alpha))
for alpha in [0.5, 0.75, 1.0, 1.25, 1.50]]),
("ContrastNormalization\n(per channel)",
[("alpha=(%.2f, %.2f)" % (
alphas[0],
alphas[1],
), iaa.ContrastNormalization(alpha=alphas, per_channel=True))
for alphas in [(0.4, 0.6), (0.65,
0.85), (0.9, 1.1), (1.15,
1.35), (1.4, 1.6)]]),
("Grayscale", [("alpha=%.1f" % (alpha, ), iaa.Grayscale(alpha=alpha))
for alpha in [0.0, 0.25, 0.5, 0.75, 1.0]]),
("PiecewiseAffine", [("scale=%.3f" % (scale, ),
iaa.PiecewiseAffine(scale=scale))
for scale in [0.015, 0.03, 0.045, 0.06, 0.075]]),
("Affine: Scale", [("%.1fx" % (scale, ), iaa.Affine(scale=scale))
for scale in [0.1, 0.5, 1.0, 1.5, 1.9]]),
("Affine: Translate",
[("x=%d y=%d" % (x, y), iaa.Affine(translate_px={
"x": x,
"y": y
}))
for x, y in [(-32, -16), (-16, -32), (-16, -8), (16, 8), (16, 32)]]),
("Affine: Rotate", [("%d deg" % (rotate, ), iaa.Affine(rotate=rotate))
for rotate in [-90, -45, 0, 45, 90]]),
("Affine: Shear", [("%d deg" % (shear, ), iaa.Affine(shear=shear))
for shear in [-45, -25, 0, 25, 45]]),
("Affine: Modes",
[(mode, iaa.Affine(translate_px=-32, mode=mode))
for mode in ["constant", "edge", "symmetric", "reflect", "wrap"]]),
("Affine: cval", [("%d" % (int(cval * 255), ),
iaa.Affine(translate_px=-32,
cval=int(cval * 255),
mode="constant"))
for cval in [0.0, 0.25, 0.5, 0.75, 1.0]]),
("Affine: all", [("",
iaa.Affine(scale={
"x": (0.5, 1.5),
"y": (0.5, 1.5)
},
translate_px={
"x": (-32, 32),
"y": (-32, 32)
},
rotate=(-45, 45),
shear=(-32, 32),
mode=ia.ALL,
cval=(0.0, 1.0))) for _ in sm.xrange(5)]),
("ElasticTransformation\n(sigma=0.2)",
[("alpha=%.1f" % (alpha, ),
iaa.ElasticTransformation(alpha=alpha, sigma=0.2))
for alpha in [0.1, 0.5, 1.0, 3.0, 9.0]])
]
print("[draw_per_augmenter_images] Augmenting...")
rows = []
for (row_name, augmenters) in rows_augmenters:
row_images = []
row_keypoints = []
row_titles = []
for img_title, augmenter in augmenters:
aug_det = augmenter.to_deterministic()
row_images.append(aug_det.augment_image(image))
row_keypoints.append(aug_det.augment_keypoints(keypoints)[0])
row_titles.append(img_title)
rows.append((row_name, row_images, row_keypoints, row_titles))
# matplotlib drawin routine
"""
print("[draw_per_augmenter_images] Plotting...")
width = 8
height = int(1.5 * len(rows_augmenters))
fig = plt.figure(figsize=(width, height))
grid_rows = len(rows)
grid_cols = 1 + 5
gs = gridspec.GridSpec(grid_rows, grid_cols, width_ratios=[2, 1, 1, 1, 1, 1])
axes = []
for i in sm.xrange(grid_rows):
axes.append([plt.subplot(gs[i, col_idx]) for col_idx in sm.xrange(grid_cols)])
fig.tight_layout()
#fig.subplots_adjust(bottom=0.2 / grid_rows, hspace=0.22)
#fig.subplots_adjust(wspace=0.005, hspace=0.425, bottom=0.02)
fig.subplots_adjust(wspace=0.005, hspace=0.005, bottom=0.02)
for row_idx, (row_name, row_images, row_keypoints, row_titles) in enumerate(rows):
axes_row = axes[row_idx]
for col_idx in sm.xrange(grid_cols):
ax = axes_row[col_idx]
ax.cla()
ax.axis("off")
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
if col_idx == 0:
ax.text(0, 0.5, row_name, color="black")
else:
cell_image = row_images[col_idx-1]
cell_keypoints = row_keypoints[col_idx-1]
cell_image_kp = cell_keypoints.draw_on_image(cell_image, size=5)
ax.imshow(cell_image_kp)
x = 0
y = 145
#ax.text(x, y, row_titles[col_idx-1], color="black", backgroundcolor="white", fontsize=6)
ax.text(x, y, row_titles[col_idx-1], color="black", fontsize=7)
fig.savefig("examples.jpg", bbox_inches="tight")
#plt.show()
"""
# simpler and faster drawing routine
output_image = ExamplesImage(128, 128, 128 + 64, 32)
for (row_name, row_images, row_keypoints, row_titles) in rows:
row_images_kps = []
for image, keypoints in zip(row_images, row_keypoints):
row_images_kps.append(keypoints.draw_on_image(image, size=5))
output_image.add_row(row_name, row_images_kps, row_titles)
misc.imsave("examples.jpg", output_image.draw())
def test_alpha_1_lightness_3(self):
aug = iaa.Sharpen(alpha=1.0, lightness=3)
observed = aug.augment_image(self.base_img)
expected = self._compute_sharpened_base_img(1.0 * 3, self.m)
assert np.allclose(observed, expected)
def draw_single_sequential_images():
image = misc.imresize(
ndimage.imread("quokka.jpg")[0:643, 0:643], (128, 128))
"""
rarely = lambda aug: iaa.Sometimes(0.1, aug)
sometimes = lambda aug: iaa.Sometimes(0.25, aug)
often = lambda aug: iaa.Sometimes(0.5, aug)
seq = iaa.Sequential([
iaa.Fliplr(0.5), # horizontally flip 50% of all images
iaa.Flipud(0.5), # vertically flip 50% of all images
rarely(iaa.Superpixels(p_replace=(0, 1.0), n_segments=(20, 200))), # convert images into their superpixel representation
often(iaa.Crop(percent=(0, 0.1))), # crop images by 0-10% of their height/width
rarely(iaa.GaussianBlur((0, 3.0))), # blur images with a sigma between 0 and 3.0
rarely(iaa.AverageBlur(k=(2, 7))), # blur image using local means with kernel sizes between 2 and 7
rarely(iaa.MedianBlur(k=(3, 11))), # blur image using local medians with kernel sizes between 2 and 7
sometimes(iaa.Sharpen(alpha=(0, 1.0), lightness=(0.75, 1.5))), # sharpen images
sometimes(iaa.Emboss(alpha=(0, 1.0), strength=(0, 2.0))), # emboss images
# search either for all edges or for directed edges
rarely(iaa.Sometimes(0.5,
iaa.EdgeDetect(alpha=(0, 0.7)),
iaa.DirectedEdgeDetect(alpha=(0, 0.7), direction=(0.0, 1.0)),
)),
often(iaa.AdditiveGaussianNoise(loc=0, scale=(0.0, 0.05*255), per_channel=0.5)), # add gaussian noise to images
often(iaa.Dropout((0.0, 0.1), per_channel=0.5)), # randomly remove up to 10% of the pixels
often(iaa.CoarseDropout((0.0, 0.05), size_percent=(0.02, 0.25), per_channel=0.5)),
rarely(iaa.Invert(0.25, per_channel=True)), # invert color channels
often(iaa.Add((-10, 10), per_channel=0.5)), # change brightness of images (by -10 to 10 of original value)
often(iaa.Multiply((0.5, 1.5), per_channel=0.5)), # change brightness of images (50-150% of original value)
often(iaa.ContrastNormalization((0.5, 2.0), per_channel=0.5)), # improve or worsen the contrast
sometimes(iaa.Grayscale(alpha=(0.0, 1.0))),
often(iaa.Affine(
scale={"x": (0.8, 1.2), "y": (0.8, 1.2)}, # scale images to 80-120% of their size, individually per axis
translate_percent={"x": (-0.2, 0.2), "y": (-0.2, 0.2)}, # translate by -20 to +20 percent (per axis)
rotate=(-45, 45), # rotate by -45 to +45 degrees
shear=(-16, 16), # shear by -16 to +16 degrees
order=[0, 1], # use nearest neighbour or bilinear interpolation (fast)
cval=(0, 255), # if mode is constant, use a cval between 0 and 255
mode=ia.ALL # use any of scikit-image's warping modes (see 2nd image from the top for examples)
)),
rarely(iaa.ElasticTransformation(alpha=(0.5, 3.5), sigma=0.25)) # move pixels locally around (with random strengths)
],
random_order=True
)
"""
sometimes = lambda aug: iaa.Sometimes(0.5, aug)
seq = iaa.Sequential(
[
# apply the following augmenters to most images
iaa.Fliplr(0.5), # horizontally flip 50% of all images
iaa.Flipud(0.2), # vertically flip 20% of all images
sometimes(iaa.Crop(
percent=(0,
0.1))), # crop images by 0-10% of their height/width
sometimes(
iaa.Affine(
scale={
"x": (0.8, 1.2),
"y": (0.8, 1.2)
}, # scale images to 80-120% of their size, individually per axis
translate_percent={
"x": (-0.2, 0.2),
"y": (-0.2, 0.2)
}, # translate by -20 to +20 percent (per axis)
rotate=(-45, 45), # rotate by -45 to +45 degrees
shear=(-16, 16), # shear by -16 to +16 degrees
order=[
0,
1
], # use nearest neighbour or bilinear interpolation (fast)
cval=(
0, 255
), # if mode is constant, use a cval between 0 and 255
mode=ia.
ALL # use any of scikit-image's warping modes (see 2nd image from the top for examples)
)),
# execute 0 to 5 of the following (less important) augmenters per image
# don't execute all of them, as that would often be way too strong
iaa.SomeOf(
(0, 5),
[
sometimes(
iaa.Superpixels(p_replace=(0, 1.0),
n_segments=(20, 200))
), # convert images into their superpixel representation
iaa.OneOf([
iaa.GaussianBlur(
(0, 3.0
)), # blur images with a sigma between 0 and 3.0
iaa.AverageBlur(
k=(2, 7)
), # blur image using local means with kernel sizes between 2 and 7
iaa.MedianBlur(
k=(3, 11)
), # blur image using local medians with kernel sizes between 2 and 7
]),
iaa.Sharpen(alpha=(0, 1.0),
lightness=(0.75, 1.5)), # sharpen images
iaa.Emboss(alpha=(0, 1.0),
strength=(0, 2.0)), # emboss images
# search either for all edges or for directed edges
sometimes(
iaa.OneOf([
iaa.EdgeDetect(alpha=(0, 0.7)),
iaa.DirectedEdgeDetect(alpha=(0, 0.7),
direction=(0.0, 1.0)),
])),
iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, 0.05 * 255),
per_channel=0.5), # add gaussian noise to images
iaa.OneOf([
iaa.Dropout(
(0.01, 0.1), per_channel=0.5
), # randomly remove up to 10% of the pixels
iaa.CoarseDropout((0.03, 0.15),
size_percent=(0.02, 0.05),
per_channel=0.2),
]),
iaa.Invert(0.05,
per_channel=True), # invert color channels
iaa.Add(
(-10, 10), per_channel=0.5
), # change brightness of images (by -10 to 10 of original value)
iaa.Multiply(
(0.5, 1.5), per_channel=0.5
), # change brightness of images (50-150% of original value)
iaa.ContrastNormalization(
(0.5, 2.0),
per_channel=0.5), # improve or worsen the contrast
iaa.Grayscale(alpha=(0.0, 1.0)),
sometimes(
iaa.ElasticTransformation(alpha=(0.5, 3.5), sigma=0.25)
), # move pixels locally around (with random strengths)
sometimes(iaa.PiecewiseAffine(scale=(
0.01,
0.05))) # sometimes move parts of the image around
],
random_order=True)
],
random_order=True)
grid = seq.draw_grid(image, cols=8, rows=8)
misc.imsave("examples_grid.jpg", grid)
def processor(self):
return iaa.Sharpen(self.alpha, self.lightness)
# Change images to grayscale and overlay them with the original image by varying strengths,
# effectively removing alpha_lo to alpha_hi of the color:
"Grayscale": lambda alpha_lo, alpha_hi: iaa.Grayscale(alpha=(alpha_lo, alpha_hi)),
# Blur each image with a gaussian kernel with a sigma between sigma_lo and sigma_hi:
"Gaussian_Blur": lambda sigma_lo, sigma_hi: iaa.GaussianBlur(sigma=(sigma_lo, sigma_hi)),
# Blur each image using a mean over neighbourhoods that have random sizes,
# which can vary between h_lo and h_hi in height and w_lo and w_hi in width:
"Average_Blur": lambda h_lo, h_hi, w_lo, w_hi: iaa.AverageBlur(k=((h_lo, h_hi), (w_lo, w_hi))),
# Blur each image using a median over neighbourhoods that have a random size between lo x lo and hi x hi:
"Median_Blur": lambda lo, hi: iaa.MedianBlur(k=(lo, hi)),
# Sharpen an image, then overlay the results with the original using an alpha between alpha_lo and alpha_hi:
"Sharpen": lambda alpha_lo, alpha_hi, lightness_lo, lightness_hi: iaa.Sharpen
(alpha=(alpha_lo, alpha_hi), lightness=(lightness_lo, lightness_hi)),
# Emboss an image, then overlay the results with the original using an alpha between alpha_lo and alpha_hi:
"Emboss": lambda alpha_lo, alpha_hi, strength_lo, strength_hi:
iaa.Emboss(alpha=(alpha_lo, alpha_hi), strength=(strength_lo, strength_hi)),
# Detect edges in images, turning them into black and white images and
# then overlay these with the original images using random alphas between alpha_lo and alpha_hi:
"Detect_Edges": lambda alpha_lo, alpha_hi: iaa.EdgeDetect(alpha=(alpha_lo, alpha_hi)),
# Detect edges having random directions between dir_lo and dir_hi (i.e (0.0, 1.0) = 0 to 360 degrees) in
# images, turning the images into black and white versions and then overlay these with the original images
# using random alphas between alpha_lo and alpha_hi:
"Directed_edge_Detect": lambda alpha_lo, alpha_hi, dir_lo, dir_hi:
iaa.DirectedEdgeDetect(alpha=(alpha_lo, alpha_hi), direction=(dir_lo, dir_hi)),
def _get_image_blob(roidb, im_scale, augment_en=False, mode='train'):
"""Builds an input blob from the images in the roidb at the specified
scales.
"""
num_images = len(roidb)
processed_ims = []
im_infos = []
for i in range(num_images):
#TODO: Probably can remove this conditional branch by just providing the filename instead of the image itself?
if(mode == 'test'):
#assert augment_en is False
im = cv2.imread(roidb[i])
local_roidb = None
else:
im = cv2.imread(roidb[i]['filename'])
local_roidb = deepcopy(roidb)
img_arr = im
mean = 0
sigma = 2
#scale
if(augment_en and mode != 'test'):
#print('augmenting image {}'.format(roidb[i]['filename']))
#shape 0 -> height
#shape 1 -> width
np.random.seed(int.from_bytes(os.urandom(4), sys.byteorder))
flip_frame = np.random.choice([True,False],p=[0.5,0.5])
if(local_roidb[i]['flipped'] is True):
print('something wrong has happened')
if(flip_frame):
img_arr = img_arr[:, ::-1, :]
local_roidb[i]['flipped'] = True
oldx1 = local_roidb[i]['boxes'][:, 0].copy()
oldx2 = local_roidb[i]['boxes'][:, 2].copy()
local_roidb[i]['boxes'][:, 0] = im.shape[1] - oldx2 - 1
local_roidb[i]['boxes'][:, 2] = im.shape[1] - oldx1 - 1
else:
local_roidb[i]['flipped'] = False
#iaa.Sometimes(0.5,(iaa.CropAndPad(
# percent=(0, 0.1),
# pad_mode='constant',
# pad_cval=(0, 255),
# keep_size=True
#))),
seq = iaa.Sequential(
[
iaa.SomeOf((0, 2),[
iaa.SomeOf((0,1),([
iaa.GaussianBlur((0.5, 2.5)), # blur images with a sigma between 0 and 3.0
iaa.AverageBlur(k=(1, 3)), # blur image using local means with kernel sizes between 2 and 7
iaa.MedianBlur(k=(1, 3)), # blur image using local medians with kernel sizes between 2 and 7
iaa.Sharpen(alpha=(0, 1.0), lightness=(0.75, 1.5))
])),
iaa.AdditiveGaussianNoise(
loc=0,
scale=(0.0, 0.1*255),
per_channel=True),
iaa.AddToHueAndSaturation((-5, 5),per_channel=True), # change hue and saturation
], random_order=True),
iaa.Sometimes(0.3,(iaa.Affine(
scale={"x": (0.9, 1.2), "y": (0.9, 1.2)}, # scale images to 80-120% of their size, individually per axis
translate_percent={"x": (-0.05, 0.05), "y": (-0.05, 0.05)}, # translate by -20 to +20 percent (per axis)
order=[0, 1], # use nearest neighbour or bilinear interpolation (fast)
cval=(0, 255), # if mode is constant, use a cval between 0 and 255
shear=(-0.05, 0.05),
mode='constant' # use any of scikit-image's warping modes (see 2nd image from the top for examples)
))),
iaa.Sometimes(0.25,iaa.Dropout((0.01, 0.05), per_channel=0.5))
#iaa.SomeOf((0,1),[
# iaa.Sometimes(0.25,iaa.Multiply((0.5, 1.5), per_channel=0.5)),
# iaa.Invert(0.05, per_channel=True)
#]),
#iaa.OneOf([
#iaa.Sometimes(0.5,iaa.ElasticTransformation(alpha=(0.5, 3.5), sigma=0.25)),
# iaa.PiecewiseAffine(scale=(0.01, 0.05))
#]),
], random_order=False
)
images_aug, bboxes_aug = seq(images=[img_arr],bounding_boxes=[local_roidb[i]['boxes']])
img_arr = images_aug[0]
local_roidb[i]['boxes'] = bboxes_aug[0]
orig_height = img_arr.shape[0]
orig_width = img_arr.shape[1]
img_arr = np.minimum(img_arr,255)
img_arr = np.maximum(img_arr,0)
img_arr = img_arr.astype('uint8')
#if(down_shift < 0):
# img_arr = np.pad(img_arr,((0,abs(down_shift)), (0,0), (0,0)), mode='constant',constant_values=(127))[abs(down_shift):,:,:]
#elif(down_shift > 0):
# img_arr = np.pad(img_arr,((abs(down_shift),0), (0,0), (0,0)), mode='constant',constant_values=(127))[:-down_shift,:,:]
#if(right_shift < 0):
# img_arr = np.pad(img_arr,((0,0), (0,abs(right_shift)), (0,0)), mode='constant',constant_values=(127))[:,abs(right_shift):,:]
#elif(right_shift > 0):
# img_arr = np.pad(img_arr,((0,0), (abs(right_shift),0), (0,0)), mode='constant',constant_values=(127))[:,:-right_shift,:]
for j, roi in enumerate(local_roidb[i]['boxes']):
#boxes[ix, :] = [x1, y1, x2, y2]
orig = roi
h = roi[3] - roi[1]
w = roi[2] - roi[0]
roi[0] = np.minimum(np.maximum(roi[0],0),orig_width-1)
roi[2] = np.minimum(np.maximum(roi[2],0),orig_width-1)
roi[1] = np.minimum(np.maximum(roi[1],0),orig_height-1)
roi[3] = np.minimum(np.maximum(roi[3],0),orig_height-1)
#TODO: magic number
if(roi[3] - roi[1] < 2):
#print('removing box y0 {} y1 {}'.format(roi[1],roi[3]))
local_roidb[i]['ignore'][j] = True
#TODO: magic number
if(roi[2] - roi[0] < 2):
#print('removing box x0 {} x1 {}'.format(roi[0],roi[2]))
local_roidb[i]['ignore'][j] = True
wc = roi[2] - roi[0]
hc = roi[3] - roi[1]
if(h != 0 and hc/h < 0.1):
local_roidb[i]['ignore'][j] = True
elif(w != 0 and wc/w < 0.1):
local_roidb[i]['ignore'][j] = True
if(local_roidb[i]['ignore'][j] is False and roi[2] < roi[0]):
print('x2 is smaller than x1')
if(local_roidb[i]['ignore'][j] is False and roi[3] < roi[1]):
print('y2 is smaller than y1')
if(local_roidb[i]['ignore'][j] is False and wc > w):
print('new x is larger than old x diff')
if(local_roidb[i]['ignore'][j] is False and hc > h):
print('new y diff is larger than old y diff')
im = img_arr
elif(augment_en and mode == 'test'):
seq = iaa.Sequential(
[
#iaa.weather.Rain(speed=(0.15,0.25))
#iaa.Fog()
iaa.imgcorruptlike.Spatter(severity=5)
#iaa.Dropout((0.6))
#iaa.GaussianBlur(sigma=(4.0,4.0))
], random_order=False)
images_aug = seq(images=[img_arr])
img_arr = images_aug[0]
orig_height = img_arr.shape[0]
orig_width = img_arr.shape[1]
img_arr = np.minimum(img_arr,255)
img_arr = np.maximum(img_arr,0)
img_arr = img_arr.astype('uint8')
im = img_arr
#minibatch(im,local_roidb[0])
#draw_and_save_minibatch(im[:,:,cfg.PIXEL_ARRANGE_BGR],roidb[i])
#TODO: Move scaling to be before imgaug, to save time
im = prep_im_for_blob(im, cfg.PIXEL_MEANS, cfg.PIXEL_STDDEVS, cfg.PIXEL_ARRANGE, im_scale)
#x_min, x_max, y_min, y_max, scale
info = np.array([0, im.shape[1], 0, im.shape[0], 0, 0, im_scale], dtype=np.float32)
im_infos.append(info)
processed_ims.append(im)
# Create a blob to hold the input images
blob = im_list_to_blob(processed_ims)
return im_infos, blob, local_roidb
translate_percent={
"x": (-0.02, 0.02),
"y": (-0.04, 0.04)
},
rotate=(-2, 2),
shear=(-1, 1),
),
iaa.PiecewiseAffine(scale=(0.001, 0.025)),
]),
iaa.OneOf([ ## brightness or contrast
iaa.Multiply((0.9, 1.1)),
iaa.ContrastNormalization((0.9, 1.1)),
]),
iaa.OneOf([ ## blur or sharpen
iaa.GaussianBlur(sigma=(0.0, 0.1)),
iaa.Sharpen(alpha=(0.0, 0.1)),
]),
])
# test on the same image as above
imggrid = augmentation.draw_grid(image[:, :, 0], cols=5, rows=2)
plt.figure(figsize=(30, 12))
_ = plt.imshow(imggrid[:, :, 0], cmap='gray')
model = modellib.MaskRCNN(mode='training', config=config, model_dir=ROOT_DIR)
# Exclude the last layers because they require a matching
# number of classes
model.load_weights(COCO_WEIGHTS_PATH,
by_name=True,
exclude=[
def _apply_aug_all():
""" Load image augmentation model """
sometimes = (lambda aug: iaa.Sometimes(0.5, aug))
return iaa.Sequential(
[
# apply the following augmenters to most images
iaa.Fliplr(0.5), # horizontally flip 50% of all images
iaa.Flipud(0.2), # vertically flip 20% of all images
# crop images by -5% to 10% of their height/width
sometimes(
iaa.CropAndPad(percent=(-0.05, 0.1),
pad_mode='constant',
pad_cval=(0, 255))),
sometimes(
iaa.Affine(
# scale images to 80-120% of their size, individually per axis
scale={
"x": (0.8, 1.2),
"y": (0.8, 1.2)
},
# translate by -20 to +20 percent (per axis)
translate_percent={
"x": (-0.2, 0.2),
"y": (-0.2, 0.2)
},
rotate=(-45, 45), # rotate by -45 to +45 degrees
shear=(-16, 16), # shear by -16 to +16 degrees
# use nearest neighbour or bilinear interpolation (fast)
order=[0, 1],
# if mode is constant, use a cval between 0 and 255
cval=(0, 255),
# use any of scikit-image's warping modes
# (see 2nd image from the top for examples)
mode='constant')),
# execute 0 to 5 of the following (less important) augmenters per
# image don't execute all of them, as that would often be way too
# strong
iaa.SomeOf(
(0, 5),
[
# convert images into their superpixel representation
sometimes(
iaa.Superpixels(p_replace=(0, 1.0),
n_segments=(20, 200))),
iaa.OneOf([
# blur images with a sigma between 0 and 3.0
iaa.GaussianBlur((0, 3.0)),
# blur image using local means with kernel sizes
# between 2 and 7
iaa.AverageBlur(k=(2, 7)),
# blur image using local medians with kernel sizes
# between 2 and 7
iaa.MedianBlur(k=(3, 11)),
]),
iaa.Sharpen(alpha=(0, 1.0),
lightness=(0.75, 1.5)), # sharpen images
iaa.Emboss(alpha=(0, 1.0),
strength=(0, 2.0)), # emboss images
# search either for all edges or for directed edges,
# blend the result with the original image using a blobby mask
iaa.SimplexNoiseAlpha(
iaa.OneOf([
iaa.EdgeDetect(alpha=(0.5, 1.0)),
iaa.DirectedEdgeDetect(alpha=(0.5, 1.0),
direction=(0.0, 1.0)),
])),
# add gaussian noise to images
iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, 0.05 * 255), per_channel=0.5),
iaa.OneOf([
# randomly remove up to 10% of the pixels
iaa.Dropout((0.01, 0.1), per_channel=0.5),
iaa.CoarseDropout((0.03, 0.15),
size_percent=(0.02, 0.05),
per_channel=0.2),
]),
# invert color channels
iaa.Invert(0.05, per_channel=True),
# change brightness of images (by -10 to 10 of original value)
iaa.Add((-10, 10), per_channel=0.5),
# change hue and saturation
iaa.AddToHueAndSaturation((-20, 20)),
# either change the brightness of the whole image (sometimes
# per channel) or change the brightness of subareas
iaa.OneOf([
iaa.Multiply((0.5, 1.5), per_channel=0.5),
iaa.FrequencyNoiseAlpha(
exponent=(-4, 0),
first=iaa.Multiply((0.5, 1.5), per_channel=True),
second=iaa.ContrastNormalization((0.5, 2.0)))
]),
# improve or worsen the contrast
iaa.ContrastNormalization((0.5, 2.0), per_channel=0.5),
iaa.Grayscale(alpha=(0.0, 1.0)),
# move pixels locally around (with random strengths)
sometimes(
iaa.ElasticTransformation(alpha=(0.5, 3.5), sigma=0.25)
),
# sometimes move parts of the image around
sometimes(iaa.PiecewiseAffine(scale=(0.01, 0.05))),
sometimes(iaa.PerspectiveTransform(scale=(0.01, 0.1)))
],
random_order=True)
],
random_order=True)
def apply_transform_on_an_image(image, bbox=None):
# Sometimes(0.5, ...) applies the given augmenter in 50% of all cases,
# e.g. Sometimes(0.5, GaussianBlur(0.3)) would blur roughly every second
# image.
sometimes = lambda aug: iaa.Sometimes(0.5, aug)
seq = iaa.Sequential(
[
# iaa.Multiply((1.2, 1.5)), # change brightness, doesn't affect BBs
# Small gaussian blur with random sigma between 0 and 0.5.
# But we only blur about 50% of all images.
# iaa.Sometimes(0.5,
# iaa.GaussianBlur(sigma=iap.Uniform(0.0, 1.0))
# ),
iaa.ContrastNormalization(
iap.Choice([1.0, 1.5, 3.0], p=[0.5, 0.3, 0.2])),
# Add gaussian noise.
# For 50% of all images, we sample the noise once per pixel.
# For the other 50% of all images, we sample the noise per pixel AND
# channel. This can change the color (not only brightness) of the
# pixels.
# iaa.AdditiveGaussianNoise(loc=0, scale=(0.0, 0.05 * 255), per_channel=0.5),
# Make some images brighter and some darker.
# In 20% of all cases, we sample the multiplier once per channel,
# which can end up changing the color of the images.
iaa.Multiply(iap.Positive(iap.Normal(0.0, 0.1)) + 1.0,
per_channel=0.2),
# Apply affine transformations to each image.
# - scale to 80-120% of image height/width (each axis independently)
# - translate by -20 to +20 relative to height/width (per axis)
# - rotate by -45 to +45 degrees
# - shear by -16 to +16 degrees
# - order: use nearest neighbour or bilinear interpolation (fast)
# - mode: use any available mode to fill newly created pixels
# see API or scikit-image for which modes are available
# - cval: if the mode is constant, then use a random brightness
# for the newly created pixels (e.g. sometimes black,
# sometimes white)
# iaa.Affine(
# scale={"x": (0.8, 1.2), "y": (0.8, 1.2)},
# translate_percent={"x": (-0.2, 0.2), "y": (-0.2, 0.2)},
# # translate_percent=0.1,
# rotate=iap.Normal(-45, 45),
# shear=(-8, 8),
# order=[0, 1],
# cval=(0, 255),
# # translate_px=iap.RandomSign(iap.Poisson(3)),
# # translate_px=3,
# # mode=["constant", "edge"]
# mode=ia.ALL
# ),
# iaa.AddElementwise(
# iap.Discretize(
# (iap.Beta(0.5, 0.5) * 2 - 1.0) * 64
# )
# ),
#
# Execute 0 to 5 of the following (less important) augmenters per
# image. Don't execute all of them, as that would often be way too
# strong.
#
iaa.SomeOf(
(0, 5),
[
# Convert some images into their superpixel representation,
# sample between 20 and 200 superpixels per image, but do
# not replace all superpixels with their average, only
# some of them (p_replace).
# sometimes(
# iaa.Superpixels(
# p_replace=(0, 0.5),
# n_segments=(1, 4)
# )
# ),
# Blur each image with varying strength using
# gaussian blur (sigma between 0 and 3.0),
# average/uniform blur (kernel size between 2x2 and 7x7)
# median blur (kernel size between 3x3 and 11x11).
iaa.OneOf([
iaa.GaussianBlur((0, 0.3)),
iaa.AverageBlur(k=(1, 2)),
iaa.MedianBlur(k=(1, 3)),
]),
# iaa.OneOf([
# iaa.GaussianBlur((0, 3.0)),
# iaa.AverageBlur(k=(2, 7)),
# iaa.MedianBlur(k=(3, 11)),
# ]),
# Sharpen each image, overlay the result with the original
# image using an alpha between 0 (no sharpening) and 1
# (full sharpening effect).
iaa.Sharpen(alpha=(0, 0.01), lightness=(0, 0.01)),
# Same as sharpen, but for an embossing effect.
iaa.Emboss(alpha=(0, 0.01), strength=(0, 0.01)),
# Search in some images either for all edges or for
# directed edges. These edges are then marked in a black
# and white image and overlayed with the original image
# using an alpha of 0 to 0.7.
sometimes(
iaa.OneOf([
iaa.EdgeDetect(alpha=(0, 0.005)),
iaa.DirectedEdgeDetect(alpha=(0, 0.001),
direction=(0.0, 0.001)),
])),
# Add gaussian noise to some images.
# In 50% of these cases, the noise is randomly sampled per
# channel and pixel.
# In the other 50% of all cases it is sampled once per
# pixel (i.e. brightness change).
# iaa.AdditiveGaussianNoise(
# loc=0, scale=(0.0, 0.001 * 255), per_channel=0.5
# ),
# Either drop randomly 1 to 10% of all pixels (i.e. set
# them to black) or drop them on an image with 2-5% percent
# of the original size, leading to large dropped
# rectangles.
# iaa.OneOf([
# iaa.Dropout((0, 0.05), per_channel=0.5),
# iaa.CoarseDropout(
# (0, 0.01), size_percent=(0.1, 0.2),
# per_channel=0.2
# ),
# ]),
# Invert each image's channel with 5% probability.
# This sets each pixel value v to 255-v.
iaa.Invert(0.1, per_channel=True), # invert color channels
# Add a value of -10 to 10 to each pixel.
#iaa.Add((-40, 40), per_channel=0.5),
# Change brightness of images (50-150% of original value).
# iaa.Multiply((0.5, 1.5), per_channel=0.5),
# Improve or worsen the contrast of images.
# iaa.contrast.LinearContrast((0.5, 2.0), per_channel=0.5),
# Convert each image to grayscale and then overlay the
# result with the original with random alpha. I.e. remove
# colors with varying strengths.
# iaa.Grayscale(alpha=(0.0, 1.0)),
# In some images move pixels locally around (with random
# strengths).
# sometimes(
# iaa.ElasticTransformation(alpha=(0.5, 1.5), sigma=0.25)
# ),
# In some images distort local areas with varying strength.
# sometimes(iaa.PiecewiseAffine(scale=(0.01, 0.05)))
],
# do all of the above augmentations in random order
random_order=True)
],
random_order=True)
# Augment BBs and images.
aug_image = seq(image=image)
return aug_image
augmentation = iaa.Sequential(
[
# apply the following augmenters to most images
iaa.Fliplr(0.33), # horizontally/vertically flip 50% of all images
iaa.Flipud(0.33),
iaa.Sometimes(0.33, iaa.Affine(rotate=(-180, 180))),
# execute 0 to 5 of the following (less important) augmenters per image
# don't execute all of them, as that would often be way too strong
iaa.SomeOf((0, 5),
[
iaa.OneOf([
iaa.GaussianBlur((0, 3.0)), # blur images with a sigma between 0 and 3.0
iaa.AverageBlur(k=(2, 7)), # blur image using local means with kernel sizes between 2 and 7
iaa.MedianBlur(k=(3, 11)), # blur image using local medians with kernel sizes between 2 and 7
]),
iaa.Sharpen(alpha=(0, 1.0), lightness=(0.75, 1.5)), # sharpen images
iaa.CropAndPad(percent=(-0.25, 0)),
iaa.AdditiveGaussianNoise(loc=0, scale=(0.0, 0.05*255), per_channel=0.5), # add gaussian noise to images
iaa.Add((-10, 10), per_channel=0.5), # change brightness of images (by -10 to 10 of original value)
iaa.AddToHueAndSaturation((-20, 20)), # change hue and saturation
iaa.LinearContrast((0.5, 2.0), per_channel=0.5), # improve or worsen the contrast
],
random_order=True
)
],
random_order=True
)
#augmentation = iaa.Sometimes(0.75, iaa.SomeOf((1, None), [
# iaa.Fliplr(0.5),
# iaa.Flipud(0.5),
from imgaug import augmenters as iaa
aug = [
iaa.LinearContrast(alpha=2),
iaa.SigmoidContrast(gain=10),
iaa.GammaContrast(gamma=2),
iaa.CLAHE(clip_limit=(1, 5)),
iaa.Grayscale(alpha=1.0),
iaa.AddToHueAndSaturation((-20, 20), per_channel=True),
iaa.BilateralBlur(d=6),
iaa.MotionBlur(k=7),
iaa.MedianBlur(k=3),
iaa.AverageBlur(k=3),
iaa.AdditiveGaussianNoise(loc=0.8, scale=(0.01, 0.08 * 255)),
iaa.ContrastNormalization((0.3, 1.5)),
iaa.Sharpen(alpha=0, lightness=1)
]
class Dataset(object):
def __init__(self, args):
self.args = args
self.char = json.load(open(self.args.char, mode='r'))
def get_file(self, data_path):
temp = []
txts = os.listdir(data_path + 'labels')
for txt in txts:
with open(data_path + 'labels/' + txt, 'r', encoding='utf-8') as f:
lines = f.readlines()
for line in lines:
def __init__(self, images, config, shuffle=True, jitter=True, norm=None):
self.generator = None
self.images = images
self.config = config
self.shuffle = shuffle
self.jitter = jitter
self.norm = norm
self.anchors = [
BoundBox(0, 0, config['ANCHORS'][2 * i],
config['ANCHORS'][2 * i + 1])
for i in range(int(len(config['ANCHORS']) // 2))
]
### augmentors by https://github.com/aleju/imgaug
sometimes = lambda aug: iaa.Sometimes(0.5, aug)
# Define our sequence of augmentation steps that will be applied to every image
# All augmenters with per_channel=0.5 will sample one value _per image_
# in 50% of all cases. In all other cases they will sample new values
# _per channel_.
self.aug_pipe = iaa.Sequential(
[
# apply the following augmenters to most images
#iaa.Fliplr(0.5), # horizontally flip 50% of all images
#iaa.Flipud(0.2), # vertically flip 20% of all images
#sometimes(iaa.Crop(percent=(0, 0.1))), # crop images by 0-10% of their height/width
sometimes(
iaa.Affine(
#scale={"x": (0.8, 1.2), "y": (0.8, 1.2)}, # scale images to 80-120% of their size, individually per axis
#translate_percent={"x": (-0.2, 0.2), "y": (-0.2, 0.2)}, # translate by -20 to +20 percent (per axis)
#rotate=(-5, 5), # rotate by -45 to +45 degrees
#shear=(-5, 5), # shear by -16 to +16 degrees
#order=[0, 1], # use nearest neighbour or bilinear interpolation (fast)
#cval=(0, 255), # if mode is constant, use a cval between 0 and 255
#mode=ia.ALL # use any of scikit-image's warping modes (see 2nd image from the top for examples)
)),
# execute 0 to 5 of the following (less important) augmenters per image
# don't execute all of them, as that would often be way too strong
iaa.SomeOf(
(0, 5),
[
#sometimes(iaa.Superpixels(p_replace=(0, 1.0), n_segments=(20, 200))), # convert images into their superpixel representation
iaa.OneOf([
iaa.GaussianBlur(
(0, 3.0)
), # blur images with a sigma between 0 and 3.0
iaa.AverageBlur(
k=(2, 7)
), # blur image using local means with kernel sizes between 2 and 7
iaa.MedianBlur(
k=(3, 11)
), # blur image using local medians with kernel sizes between 2 and 7
]),
iaa.Sharpen(alpha=(0, 1.0),
lightness=(0.75, 1.5)), # sharpen images
#iaa.Emboss(alpha=(0, 1.0), strength=(0, 2.0)), # emboss images
# search either for all edges or for directed edges
#sometimes(iaa.OneOf([
# iaa.EdgeDetect(alpha=(0, 0.7)),
# iaa.DirectedEdgeDetect(alpha=(0, 0.7), direction=(0.0, 1.0)),
#])),
iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, 0.05 * 255),
per_channel=0.5), # add gaussian noise to images
iaa.OneOf([
iaa.Dropout(
(0.01, 0.1), per_channel=0.5
), # randomly remove up to 10% of the pixels
#iaa.CoarseDropout((0.03, 0.15), size_percent=(0.02, 0.05), per_channel=0.2),
]),
#iaa.Invert(0.05, per_channel=True), # invert color channels
iaa.Add(
(-10, 10), per_channel=0.5
), # change brightness of images (by -10 to 10 of original value)
iaa.Multiply(
(0.5, 1.5), per_channel=0.5
), # change brightness of images (50-150% of original value)
iaa.ContrastNormalization(
(0.5, 2.0),
per_channel=0.5), # improve or worsen the contrast
#iaa.Grayscale(alpha=(0.0, 1.0)),
#sometimes(iaa.ElasticTransformation(alpha=(0.5, 3.5), sigma=0.25)), # move pixels locally around (with random strengths)
#sometimes(iaa.PiecewiseAffine(scale=(0.01, 0.05))) # sometimes move parts of the image around
],
random_order=True)
],
random_order=True)
if shuffle: np.random.shuffle(self.images)
def test_alpha_one(self):
aug = iaa.Sharpen(alpha=1.0, lightness=1)
observed = aug.augment_image(self.base_img)
expected = self.base_img_sharpened
assert np.allclose(observed, expected)
def __init__(self):
# Sometimes(0.5, ...) applies the given augmenter in 50% of all cases,
# e.g. Sometimes(0.5, GaussianBlur(0.3)) would blur roughly every second image.
sometimes = lambda aug: iaa.Sometimes(.90, aug)
# Define our sequence of augmentation steps that will be applied to every image
# All augmenters with per_channel=0.5 will sample one value _per image_
# in 50% of all cases. In all other cases they will sample new values
# _per channel_.
seq = iaa.Sequential(
[
sometimes(
iaa.CropAndPad(percent=(-0.03, 0.03),
pad_mode=["constant", "edge"],
pad_cval=(0, 255))),
sometimes(
iaa.Affine(
#scale={"x": (0.8, 1.2), "y": (0.8, 1.2)}, # scale images to 80-120% of their size, individually per axis
#translate_percent={"x": (-0.1, 0.1), "y": (-0.1, 0.1)}, # translate by -20 to +20 percent (per axis)
rotate=(-3, 3), # rotate by -45 to +45 degrees
shear=(-3, 3), # shear by -16 to +16 degrees
#order=[0, 1], # use nearest neighbour or bilinear interpolation (fast)
#cval=(0, 255), # if mode is constant, use a cval between 0 and 255
#mode=ia.ALL # use any of scikit-image's warping modes (see 2nd image from the top for examples)
)),
iaa.SomeOf(
(0, 3),
[
#iaa.OneOf([
# iaa.GaussianBlur((0, 0.8)), # blur images with a sigma between 0 and 3.0
#]),
iaa.GaussianBlur(
(0, 0.75)
), # blur images with a sigma between 0 and 3.0
iaa.Sharpen(alpha=(0, 0.5),
lightness=(0.75, 1.5)), # sharpen images
iaa.Emboss(alpha=(0, 0.5),
strength=(0, 1.5)), # emboss images
#iaa.SimplexNoiseAlpha(iaa.OneOf([
# iaa.EdgeDetect(alpha=(0.5, 1.0)),
# iaa.DirectedEdgeDetect(alpha=(0.5, 1.0), direction=(0.0, 1.0)),
#])),
iaa.AdditiveGaussianNoise(
loc=0, scale=(0.0, 0.05 * 255),
per_channel=0.5), # add gaussian noise to images
iaa.Dropout(
(0.01, 0.05), per_channel=0.5
), # randomly remove up to 10% of the pixels
#iaa.OneOf([
# iaa.Dropout((0.01, 0.05), per_channel=0.5), # randomly remove up to 10% of the pixels
# iaa.CoarseDropout((0.01, 0.05), size_percent=(0.02, 0.04), per_channel=0.2),
#]),
iaa.Invert(0.05,
per_channel=True), # invert color channels
iaa.Add(
(-5, 5), per_channel=0.5
), # change brightness of images (by -10 to 10 of original value)
iaa.AddToHueAndSaturation(
(-20, 20)), # change hue and saturation
#iaa.OneOf([
# iaa.Multiply((0.5, 1.5), per_channel=0.5),
# iaa.FrequencyNoiseAlpha(
# exponent=(-4, 0),
# first=iaa.Multiply((0.5, 1.5), per_channel=True),
# second=iaa.ContrastNormalization((0.5, 2.0))
# )
#]),
iaa.ContrastNormalization(
(0.5, 1.0),
per_channel=0.5), # improve or worsen the contrast
iaa.Grayscale(alpha=(0.0, 1.0)),
],
random_order=True)
],
random_order=True)
self.seq = seq
def __init__(self,
mean=(0.0, 0.0, 0.0, 0.0),
std=(1.0, 1.0, 1.0, 1.0),
crop_image_size=(224, 224)):
if isinstance(crop_image_size, int):
crop_image_size = (crop_image_size, crop_image_size)
self._mean = mean
self._std = std
self.crop_image_size = crop_image_size
self.seq = iaa.Sequential(
children=[
iaa.Sequential(
children=[
iaa.Fliplr(
p=0.5,
name="Fliplr"),
iaa.Flipud(
p=0.5,
name="Flipud"),
iaa.Sequential(
children=[
iaa.Affine(
scale={"x": (0.9, 1.1), "y": (0.9, 1.1)},
translate_percent={"x": (-0.05, 0.05), "y": (-0.05, 0.05)},
rotate=(-45, 45),
shear=(-16, 16),
order=iap.Choice([0, 1, 3], p=[0.15, 0.80, 0.05]),
mode="reflect",
name="Affine"),
iaa.Sometimes(
p=0.01,
then_list=iaa.PiecewiseAffine(
scale=(0.0, 0.01),
nb_rows=(4, 20),
nb_cols=(4, 20),
order=iap.Choice([0, 1, 3], p=[0.15, 0.80, 0.05]),
mode="reflect",
name="PiecewiseAffine"))],
random_order=True,
name="GeomTransform"),
iaa.Sequential(
children=[
iaa.Sometimes(
p=0.75,
then_list=iaa.Add(
value=(-10, 10),
per_channel=0.5,
name="Brightness")),
iaa.Sometimes(
p=0.05,
then_list=iaa.Emboss(
alpha=(0.0, 0.5),
strength=(0.5, 1.2),
name="Emboss")),
iaa.Sometimes(
p=0.1,
then_list=iaa.Sharpen(
alpha=(0.0, 0.5),
lightness=(0.5, 1.2),
name="Sharpen")),
iaa.Sometimes(
p=0.25,
then_list=iaa.ContrastNormalization(
alpha=(0.5, 1.5),
per_channel=0.5,
name="ContrastNormalization"))
],
random_order=True,
name="ColorTransform"),
iaa.Sequential(
children=[
iaa.Sometimes(
p=0.5,
then_list=iaa.AdditiveGaussianNoise(
loc=0,
scale=(0.0, 10.0),
per_channel=0.5,
name="AdditiveGaussianNoise")),
iaa.Sometimes(
p=0.1,
then_list=iaa.SaltAndPepper(
p=(0, 0.001),
per_channel=0.5,
name="SaltAndPepper"))],
random_order=True,
name="Noise"),
iaa.OneOf(
children=[
iaa.Sometimes(
p=0.05,
then_list=iaa.MedianBlur(
k=3,
name="MedianBlur")),
iaa.Sometimes(
p=0.05,
then_list=iaa.AverageBlur(
k=(2, 4),
name="AverageBlur")),
iaa.Sometimes(
p=0.5,
then_list=iaa.GaussianBlur(
sigma=(0.0, 2.0),
name="GaussianBlur"))],
name="Blur"),
],
random_order=True,
name="MainProcess")])
