Skip to content
Snippets Groups Projects
Commit 438e6978 authored by Freek Daniëls's avatar Freek Daniëls
Browse files

Update code/4. Data loader.ipynb

parent 3721c600
Branches develop-freek
No related tags found
3 merge requests!7Develop,!6Update code/4. Data loader.ipynb,!5Update code/4. Data loader.ipynb
Hide whitespace changes
Inline Side-by-side
code/4. Data loader.ipynb
+ 1
− 1
View file @ 438e6978
%% Cell type:code id: tags: %% Cell type:code id: tags:
``` python ``` python
from __future__ import print_function from __future__ import print_function
from __future__ import division from __future__ import division
import pandas as pd import pandas as pd
import numpy as np import numpy as np
import glob import glob
import os import os
import random import random
import time import time
import copy import copy
import json import json
from dataclasses import dataclass from dataclasses import dataclass
from PIL import Image from PIL import Image
from skimage import io from skimage import io
import matplotlib.pyplot as plt import matplotlib.pyplot as plt
import torch.utils.data import torch.utils.data
import torch import torch
import torchvision.transforms.functional as TF import torchvision.transforms.functional as TF
from IPython.display import clear_output from IPython.display import clear_output
``` ```
%% Cell type:code id: tags: %% Cell type:code id: tags:
``` python ``` python
%run "0. Config.ipynb" %run "0. Config.ipynb"
%run "1. Utils.ipynb" %run "1. Utils.ipynb"
``` ```
%% Cell type:code id: tags: %% Cell type:code id: tags:
``` python ``` python
class SeagrassDataset(torch.utils.data.Dataset): class SeagrassDataset(torch.utils.data.Dataset):
def __init__(self, path_to_images, path_to_masks, path_to_configurations, path_to_files, transforms=False): def __init__(self, path_to_images, path_to_masks, path_to_configurations, path_to_files, transforms=False):
self.path_to_images = path_to_images self.path_to_images = path_to_images
self.path_to_masks = path_to_masks self.path_to_masks = path_to_masks
self.annotations = [] self.annotations = []
files = pd.read_csv(path_to_files,usecols=["file"]) files = pd.read_csv(path_to_files,usecols=["file"])
for row in files['file']: for row in files['file']:
configurations = path_to_configurations + row + ".json" configurations = path_to_configurations + row + ".json"
with open(configurations) as json_file: with open(configurations) as json_file:
data = json.load(json_file) data = json.load(json_file)
name = data['name'] name = data['name']
for configuration in data['configurations']: for configuration in data['configurations']:
annotation = {} annotation = {}
annotation['name'] = name annotation['name'] = name
annotation['configuration'] = configuration annotation['configuration'] = configuration
self.annotations.append(annotation) self.annotations.append(annotation)
self.transforms = transforms self.transforms = transforms
def __getitem__(self, idx): def __getitem__(self, idx):
annotation = self.annotations[idx] annotation = self.annotations[idx]
name = annotation['name'] name = annotation['name']
configuration = annotation["configuration"] configuration = annotation["configuration"]
configuration_mask = configuration["mask"] configuration_mask = configuration["mask"]
mask_left = configuration_mask['left'] mask_left = configuration_mask['left']
mask_right = configuration_mask['right'] mask_right = configuration_mask['right']
mask_top = configuration_mask['top'] mask_top = configuration_mask['top']
mask_bottom = configuration_mask['bottom'] mask_bottom = configuration_mask['bottom']
image = io.imread(self.path_to_images + name) image = io.imread(self.path_to_images + name)
mask = io.imread(self.path_to_masks + name) mask = io.imread(self.path_to_masks + name)
if self.transforms: if self.transforms:
image = pad_image(image,image_padding_for_augmentation) image = pad_image(image,image_padding_for_augmentation)
mask = pad_image(mask,mask_padding_for_augmentation) mask = pad_image(mask,mask_padding_for_augmentation)
image_offset = image_padding_for_augmentation image_offset = image_padding_for_augmentation
mask_offset = mask_padding_for_augmentation mask_offset = mask_padding_for_augmentation
else: else:
image = pad_image(image,image_padding) image = pad_image(image,image_padding)
image_offset = image_padding image_offset = image_padding
mask_offset = 0 mask_offset = 0
image_left = mask_left-image_offset+image_offset image_left = mask_left-image_offset+image_offset
image_top = mask_top-image_offset+image_offset image_top = mask_top-image_offset+image_offset
image_right = mask_right+image_offset+image_offset image_right = mask_right+image_offset+image_offset
image_bottom = mask_bottom+image_offset+image_offset image_bottom = mask_bottom+image_offset+image_offset
mask_left = mask_left-mask_offset+mask_offset mask_left = mask_left-mask_offset+mask_offset
mask_top = mask_top-mask_offset+mask_offset mask_top = mask_top-mask_offset+mask_offset
mask_right = mask_right+mask_offset+mask_offset mask_right = mask_right+mask_offset+mask_offset
mask_bottom = mask_bottom+mask_offset+mask_offset mask_bottom = mask_bottom+mask_offset+mask_offset
image = Image.fromarray(image).convert("RGB") image = Image.fromarray(image).convert("RGB")
mask = Image.fromarray(mask) mask = Image.fromarray(mask)
if self.transforms: if self.transforms:
image = image.crop((image_left,image_top,image_right,image_bottom)) image = image.crop((image_left,image_top,image_right,image_bottom))
mask = mask.crop((mask_left,mask_top,mask_right,mask_bottom)) mask = mask.crop((mask_left,mask_top,mask_right,mask_bottom))
if random.random() > 0.5: if random.random() > 0.5:
image = TF.hflip(image) image = TF.hflip(image)
mask = TF.hflip(mask) mask = TF.hflip(mask)
if random.random() > 0.5: if random.random() > 0.5:
image = TF.vflip(image) image = TF.vflip(image)
mask = TF.vflip(mask) mask = TF.vflip(mask)
degrees = random.random()*180 degrees = random.random()*180
image = TF.affine(image,degrees,translate=[0,0], scale=1, shear=0.0) image = TF.affine(image,degrees,translate=[0,0], scale=1, shear=0.0)
mask = TF.affine(mask,degrees,translate=[0,0], scale=1, shear=0.0) mask = TF.affine(mask,degrees,translate=[0,0], scale=1, shear=0.0)
image = TF.center_crop(image,image_size) image = TF.center_crop(image,image_size)
mask = TF.center_crop(mask,mask_size) mask = TF.center_crop(mask,mask_size)
image = TF.to_tensor(image) image = TF.to_tensor(image)
image = TF.normalize(image,mean,std) image = TF.normalize(image,mean,std)
mask = torch.tensor(np.array(mask)).long() mask = torch.tensor(np.array(mask)).long()
return image, mask, name return image, mask, name
def __len__(self): def __len__(self):
return len(self.annotations) return len(self.annotations)
``` ```
%% Cell type:code id: tags: %% Cell type:code id: tags:
``` python ``` python
datasets = {"train":SeagrassDataset(path_to_images,path_to_masks,path_to_configurations,training_file,True), datasets = {"train":SeagrassDataset(path_to_images,path_to_masks,path_to_configurations,training_file,True),
"val":SeagrassDataset(path_to_images,path_to_masks,path_to_configurations,validation_file), "val":SeagrassDataset(path_to_images,path_to_masks,path_to_configurations,validation_file),
"test":SeagrassDataset(path_to_images,path_to_masks,path_to_configurations,test_file)} "test":SeagrassDataset(path_to_images,path_to_masks,path_to_configurations,test_file)}
``` ```
%% Cell type:code id: tags: %% Cell type:code id: tags:
``` python ``` python
image,mask,name = next(iter(datasets["train"])) image,mask,name = next(iter(datasets["train"]))
image = np.transpose(np.asarray(image),axes=[1,2,0])*std+mean image = np.transpose(np.asarray(image),axes=[1,2,0])*std+mean
mask = np.asarray(mask) mask = np.asarray(mask)
plt.figure(figsize=(10,5)) plt.figure(figsize=(10,5))
plt.subplot(1, 2, 1) plt.subplot(1, 2, 1)
plt.axis('off') plt.axis('off')
plt.tight_layout() plt.tight_layout()
plt.imshow(image) plt.imshow(image)
plt.subplot(1, 2, 2) plt.subplot(1, 2, 2)
plt.axis('off') plt.axis('off')
plt.tight_layout() plt.tight_layout()
plt.imshow(get_coloured_mask(mask,labels)) plt.imshow(get_coloured_mask(mask,labels))
plt.show() plt.show()
``` ```
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment