Create image_labels.py

pre_process/GlaS/image_labels.py

@@ -0,0 +1,97 @@
+# This script is used to split the test set of the official GlaS into train-test split for our downstream
+# histopathological image segmentation
+
+from PIL import Image
+from torchvision.transforms import transforms
+import torch
+import glob
+import json
+import numpy as np
+import os
+import cv2
+from skimage import io
+import sys
+from scipy.io import loadmat, savemat
+
+
+def split_data(images, labels, ratio):
+    idxs = np.random.RandomState(2023).permutation(images.shape[0])
+    split = int(images.shape[0] * ratio)
+    split_1 = idxs[:split]
+    split_2 = idxs[split:]
+    return images[split_1], images[split_2], labels[split_1], labels[split_2]
+
+
+folder1 = './data/test_images'
+folder2 = './data/test_masks'
+
+# directories for images and labels
+TRAIN_OUT_FOLDER = './train/full_size_images'
+os.makedirs(TRAIN_OUT_FOLDER, exist_ok=True)
+TEST_OUT_FOLDER = './test/full_size_images'
+os.makedirs(TEST_OUT_FOLDER, exist_ok=True)
+
+TRAIN_LABEL_FOLDER = './train/full_size_masks'
+os.makedirs(TRAIN_LABEL_FOLDER, exist_ok=True)
+TEST_LABEL_FOLDER = './test/full_size_masks'
+os.makedirs(TEST_LABEL_FOLDER, exist_ok=True)
+
+print('===========================================================================')
+print('                          IMAGES and LABELS                                    ')
+print('===========================================================================')
+
+total_list1 = os.listdir(folder1)
+total_list1 = np.array(sorted(total_list1, key=lambda x: int(x.split('_')[-1].split('.jpg')[0])))
+
+total_list2 = os.listdir(folder2)
+total_list2 = np.array(sorted(total_list2, key=lambda x: int(x.split('_')[-1].split('.jpg')[0])))
+
+ratio = 0.8
+idxs = np.random.RandomState(2023).permutation(total_list1.shape[0])
+split = int(total_list1.shape[0] * ratio)
+split_1 = idxs[:split]
+split_2 = idxs[split:]
+train_images, test_images = total_list1[split_1], total_list1[split_2]
+train_labels, test_labels = total_list2[split_1], total_list2[split_2]
+
+i = 0
+for img_name in train_images:
+    img = cv2.imread(os.path.join(folder1, img_name))
+    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+    TRAIN_OUT_IMAGE_PATH = os.path.join(TRAIN_OUT_FOLDER, 'image_' + str(i) + '.jpg')
+    cv2.imwrite(TRAIN_OUT_IMAGE_PATH, cv2.cvtColor(img, cv2.COLOR_RGB2BGR))
+    print(img_name)
+    i += 1
+
+print('='*30)
+
+i = 0
+for label_name in train_labels:
+    mask = cv2.imread(os.path.join(folder2, label_name))
+    mask = cv2.cvtColor(mask, cv2.COLOR_BGR2GRAY)
+    TRAIN_OUT_IMAGE_PATH = os.path.join(TRAIN_LABEL_FOLDER, 'label_' + str(i) + '.jpg')
+    cv2.imwrite(TRAIN_OUT_IMAGE_PATH, mask)
+    print(label_name)
+    i += 1
+
+print('='*30)
+
+i = 0
+for img_name in test_images:
+    img = cv2.imread(os.path.join(folder1, img_name))
+    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+    TEST_OUT_IMAGE_PATH = os.path.join(TEST_OUT_FOLDER, 'image_' + str(i) + '.jpg')
+    cv2.imwrite(TEST_OUT_IMAGE_PATH, cv2.cvtColor(img, cv2.COLOR_RGB2BGR))
+    print(img_name)
+    i += 1
+
+print('='*30)
+
+i = 0
+for label_name in test_labels:
+    mask = cv2.imread(os.path.join(folder2, label_name))
+    mask = cv2.cvtColor(mask, cv2.COLOR_BGR2GRAY)
+    TEST_OUT_IMAGE_PATH = os.path.join(TEST_LABEL_FOLDER, 'label_' + str(i) + '.jpg')
+    cv2.imwrite(TEST_OUT_IMAGE_PATH, mask)
+    print(label_name)
+    i += 1