Lung_image_segmentation/helper.py at master · ViiSkor/Lung_image_segmentation · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
from keras.preprocessing.image import ImageDataGenerator
import matplotlib.pyplot as plt
import numpy as np
import random
import cv2
import os

ABBR = "CHNCXR"

def get_dataset_pathes(img_paths, mask_paths):
    without_mask = []
    dataset_images = []
    dataset_masks = []

    for image_dir in img_paths:
        mask_dir = image_dir.replace("images", "masks")
        if ABBR in mask_dir:
            mask_dir = '_'.join([mask_dir[:-4], "mask.png"])

        if mask_dir in mask_paths:
            dataset_images.append(image_dir)
            dataset_masks.append(mask_dir)
        else:
            without_mask.append(image_dir)

    return without_mask, np.array(dataset_images), np.array(dataset_masks)

def split_test_train(dataset_image_paths, dataset_mask_paths, ratio=0.8, seed=42):
  idx_lst = [i for i in range(len(dataset_image_paths))]
  random.seed(seed)
  random.shuffle(idx_lst)
  cut_idx = int(len(dataset_image_paths)*ratio)
  train_image_paths = dataset_image_paths[idx_lst[:cut_idx]]
  train_mask_paths = dataset_mask_paths[idx_lst[:cut_idx]]
  test_image_paths = dataset_image_paths[idx_lst[cut_idx:]]
  test_mask_paths = dataset_mask_paths[idx_lst[cut_idx:]]

  train_image_paths = train_image_paths.tolist()
  train_mask_paths = train_mask_paths.tolist()
  test_image_paths = test_image_paths.tolist()
  test_mask_paths = test_mask_paths.tolist()

  return train_image_paths, train_mask_paths, test_image_paths, test_mask_paths

def creat_dataset_dir(image_dir, masks_dir, image_paths, mask_paths):
  os.makedirs(image_dir)
  os.makedirs(masks_dir)

  for i, (image, mask) in enumerate(zip(image_paths, mask_paths)):
    os.rename(image, image_dir+str(i)+".png")
    os.rename(mask, masks_dir+str(i)+".png")

def preprocess(image, mask):
    image = image / 255
    mask = mask / 255
    mask[mask > 0.5] = 1
    mask[mask <= 0.5] = 0

    return image, mask

def generator(batch_size, path, target_size=(512, 512), augmentation=False, seed=42):
  dg_args = {}
  if augmentation:
    dg_args = dict(rotation_range=0.2,
                     width_shift_range=0.08,
                     height_shift_range=0.08,
                     shear_range=0.05,
                     zoom_range = 0.05,
                     horizontal_flip=True,
                     vertical_flip = False,
                     fill_mode='nearest')
  image_datagen = ImageDataGenerator(**dg_args)
  mask_datagen = ImageDataGenerator(**dg_args)

  image_generator = image_datagen.flow_from_directory(
      path,
      classes = ["images"],
      target_size = target_size,
      batch_size = batch_size,
      seed = seed)

  mask_generator = mask_datagen.flow_from_directory(
      path,
      classes = ["masks"],
      color_mode = "grayscale",
      target_size = target_size,
      batch_size = batch_size,
      seed = seed)

  for image, mask in zip(image_generator, mask_generator):
    image, mask = preprocess(image[0], mask[0])
    yield (image, mask)

def plot_data(dataset_images, dataset_masks, generator):
  fig = plt.figure(figsize=(24, 10))

  for i in range(16):
    image, mask = next(generator)
    plt.subplot(4, 8, i*2+1)
    plt.imshow(image[0, :, :, :])
    plt.axis('off')
    plt.subplot(4, 8, i*2+2)
    plt.imshow(mask[0, :, :, 0], cmap='gray')
    plt.axis('off')

  plt.show()

def plot_result(X, y, prediction, n_show):
  fig = plt.figure(figsize=(15, 12))

  for i in range(n_show):
    j = np.random.randint(len(X))
    plt.subplot(n_show, 4, i*4+1)
    plt.imshow(X[j, :, :, :])
    plt.axis('off')

    plt.subplot(n_show, 4, i*4+2)
    plt.gca().set_title('Origin mask')
    plt.imshow(y[j, :, :, 0], cmap='gray')
    plt.axis('off')

    plt.subplot(n_show, 4, i*4+3)
    plt.gca().set_title('Predicted mask')
    plt.imshow(prediction[j, :, :, 0], cmap='gray')
    plt.axis('off')

    overlay = cv2.addWeighted(y[j, :, :, 0], 0.5, prediction[j, :, :, 0], 0.5, 0)
    plt.subplot(n_show, 4, i*4+4)
    plt.gca().set_title('Overlay origin and predicted mask')
    plt.imshow(overlay)
    plt.axis('off')

  plt.show()