def main(args):
img = ecvl.ImRead(args.in_img)
augs = ecvl.SequentialAugmentationContainer([
ecvl.AugRotate([-5, 5]),
ecvl.AugMirror(.5),
ecvl.AugFlip(.5),
ecvl.AugGammaContrast([3, 5]),
ecvl.AugAdditiveLaplaceNoise([0, 0.2 * 255]),
ecvl.AugCoarseDropout([0, 0.55], [0.02, 0.1], 0.5),
ecvl.AugAdditivePoissonNoise([0, 40]),
ecvl.AugResizeDim([500, 500]),
])
ecvl.AugmentationParam.SetSeed(0)
augs.Apply(img)
print("Executing ImageToTensor")
t = ecvl.ImageToTensor(img)
t.div_(128)
t.mult_(128)
print("Executing TensorToImage")
img = ecvl.TensorToImage(t)
print("Executing TensorToView")
ecvl.TensorToView(t)
_ = ecvl.AugmentationFactory.create(AUG_TXT)
training_augs = ecvl.SequentialAugmentationContainer([
ecvl.AugRotate([-5, 5]),
ecvl.AugAdditiveLaplaceNoise([0, 0.2 * 255]),
ecvl.AugCoarseDropout([0, 0.55], [0.02, 0.1], 0),
ecvl.AugAdditivePoissonNoise([0, 40]),
ecvl.AugResizeDim([30, 30]),
])
test_augs = ecvl.SequentialAugmentationContainer([
ecvl.AugResizeDim([30, 30]),
])
ds_augs = ecvl.DatasetAugmentations([training_augs, None, test_augs])
batch_size = 64
print("Creating a DLDataset")
d = ecvl.DLDataset(args.in_ds, batch_size, ds_augs, ecvl.ColorType.GRAY)
print("Create x and y")
x = Tensor(
[batch_size, d.n_channels_, d.resize_dims_[0], d.resize_dims_[1]])
y = Tensor([batch_size, len(d.classes_)])
# Load a batch of d.batch_size_ images into x and corresponding labels
# into y. Images are resized to the dimensions specified in the
# augmentations chain
print("Executing LoadBatch on training set")
d.LoadBatch(x, y)
# Change colortype and channels
img = ecvl.TensorToImage(x)
img.colortype_ = ecvl.ColorType.GRAY
img.channels_ = "xyc"
# Switch to Test split and load a batch of images
print("Executing LoadBatch on test set")
d.SetSplit(ecvl.SplitType.test)
d.LoadBatch(x, y)
def read_slide(slide_fn, level=4):
levels = ecvl.OpenSlideGetLevels(slide_fn)
dims = [0, 0] + levels[level]
img = ecvl.OpenSlideRead(slide_fn, level, dims)
t = ecvl.ImageToTensor(img)
t_np = t.getdata()
s = t_np.shape
t_np = t_np.transpose((1, 2, 0)).reshape(
(s[1] * s[2], 3)) # Channel last and reshape
t_eval = Tensor.fromarray(t_np)
print(t_eval.getShape())
return t_eval, s
def get_mask_tissue_from_slide(self,
slide_fn,
level=2,
use_openslide=False):
if use_openslide and not no_openslide:
# Using open slide
slide = openslide.open_slide(slide_fn)
dims = slide.level_dimensions
ds = [int(i) for i in slide.level_downsamples]
x0 = 0
y0 = 0
x1 = dims[0][0]
y1 = dims[0][1]
delta_x = x1 - x0
delta_y = y1 - y0
pil_img = slide.read_region(location=(x0, y0),
level=level,
size=(delta_x // ds[level],
delta_y // ds[level]))
np_img = np.array(pil_img)
msk_pred = self.get_tissue_mask(np_img,
channel_first=False,
BGR=False)
else:
## Using ECVL
levels = ecvl.OpenSlideGetLevels(slide_fn)
dims = [0, 0] + levels[level]
img = ecvl.OpenSlideRead(slide_fn, level, dims)
t = ecvl.ImageToTensor(img)
np_img = t.getdata(
) # Tensor to numpy array (ecvl read images with channel first)
msk_pred = self.get_tissue_mask(np_img,
channel_first=True,
BGR=True)
return msk_pred
def main(args):
if not ecvl.ECVL_EDDL:
print("No EDDL support - quitting")
sys.exit(0)
img = ecvl.ImRead(args.in_img)
augs = ecvl.SequentialAugmentationContainer([
ecvl.AugCenterCrop(), # Make image square
ecvl.AugRotate([-5, 5]),
ecvl.AugMirror(.5),
ecvl.AugFlip(.5),
ecvl.AugGammaContrast([3, 5]),
ecvl.AugAdditiveLaplaceNoise([0, 0.2 * 255]),
ecvl.AugCoarseDropout([0, 0.55], [0.02, 0.1], 0.5),
ecvl.AugAdditivePoissonNoise([0, 40]),
ecvl.AugResizeDim([500, 500]),
ecvl.AugCenterCrop([224, 224]),
ecvl.AugToFloat32(255),
ecvl.AugNormalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]),
])
ecvl.AugmentationParam.SetSeed(0)
print("Applying augmentations")
augs.Apply(img)
print("Executing ImageToTensor")
t = ecvl.ImageToTensor(img)
t.div_(128)
t.mult_(128)
print("Executing TensorToImage")
img = ecvl.TensorToImage(t)
print("Executing TensorToView")
ecvl.TensorToView(t)
print("Applying augmentations (from text)")
newdeal_augs = ecvl.AugmentationFactory.create(AUG_TXT)
newdeal_augs.Apply(img)
training_augs = ecvl.SequentialAugmentationContainer([
ecvl.AugRotate([-5, 5]),
ecvl.AugAdditiveLaplaceNoise([0, 0.2 * 255]),
ecvl.AugCoarseDropout([0, 0.55], [0.02, 0.1], 0),
ecvl.AugAdditivePoissonNoise([0, 40]),
ecvl.AugResizeDim([30, 30]),
ecvl.AugToFloat32(255),
ecvl.AugNormalize(0.449, 0.226), # mean of imagenet stats
])
test_augs = ecvl.SequentialAugmentationContainer([
ecvl.AugResizeDim([30, 30]),
ecvl.AugToFloat32(255),
ecvl.AugNormalize(0.449, 0.226), # mean of imagenet stats
])
# number of augmentation containers must match number of dataset splits
ds_augs = ecvl.DatasetAugmentations([training_augs, test_augs])
batch_size = 64
print("Creating a DLDataset")
d = ecvl.DLDataset(args.in_ds, batch_size, ds_augs, ecvl.ColorType.GRAY)
print("Create x and y")
x = Tensor(
[batch_size, d.n_channels_, d.resize_dims_[0], d.resize_dims_[1]])
y = Tensor([batch_size, len(d.classes_)])
# Load a batch of d.batch_size_ images into x and corresponding labels
# into y. Images are resized to the dimensions specified in the
# augmentations chain
print("Executing LoadBatch on training set")
d.LoadBatch(x, y)
# Change colortype and channels
img = ecvl.TensorToImage(x)
img.colortype_ = ecvl.ColorType.GRAY
img.channels_ = "xyc"
# Switch to Test split and load a batch of images
print("Executing LoadBatch on test set")
d.SetSplit(ecvl.SplitType.test)
d.LoadBatch(x, y)
# Save some input images
ecvl.ImWrite("mnist_batch.png", ecvl.MakeGrid(x, 8, False))
ecvl.ImWrite("mnist_batch_normalized.png", ecvl.MakeGrid(x, 8, True))
def PneumothoraxLoadBatch(d, black_indices, m_i, b_i):
mask_indices = d.GetSplit()
len_mask_indices = len(mask_indices)
bs = d.batch_size_
names = []
images = []
labels = []
if d.current_split_ == ecvl.SplitType.training:
index = 0
elif d.current_split_ == ecvl.SplitType.validation:
index = 1
else:
index = 2
start = d.current_batch_[index] * bs
d.current_batch_ = [
elem + 1 if i == index else elem
for i, elem in enumerate(d.current_batch_)
]
expr = True
for i in range(start, start + bs):
if d.current_split_ == ecvl.SplitType.training:
# in training, check if you can take
# other black ground truth images..
if len_mask_indices * 1.25 - i > len_mask_indices - m_i:
expr = random.random() >= 0.2 and m_i < len_mask_indices
# ..otherwise, you have to take a ground truth with mask
else:
# in validation, first take all the samples with the ground
# truth with mask and then all those with the black ground truth.
expr = m_i < len_mask_indices
if expr:
index = mask_indices[m_i]
m_i += 1
else:
index = black_indices[b_i]
b_i += 1
# insert the original name of images and ground truth in case
# you want to save predictions during validation
elem = d.samples_[index]
names.append(elem.location_[0])
names.append(elem.label_path_)
# Read the image
img = elem.LoadImage(d.ctype_, False)
# Read the ground truth
gt = elem.LoadImage(d.ctype_gt_, True)
# Apply chain of augmentations to sample image
# and corresponding ground truth
d.augs_.Apply(d.current_split_, img, gt)
# Copy image into tensor (images)
images.append(ecvl.ImageToTensor(img))
# Copy label into tensor (labels)
labels.append(ecvl.ImageToTensor(gt))
return d, images, labels, names, m_i, b_i