-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset.py
43 lines (32 loc) · 1.1 KB
/
dataset.py
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
import os
import torch.utils.data as data
from PIL import Image
import torch
class ContourDataset(data.Dataset):
def __init__(self,
img_path,
skt_path,
list_path,
transformer=None):
self.img_path = img_path
self.skt_path = skt_path
with open(list_path) as f:
content = f.readlines()
self.filelist = sorted([x.strip() for x in content])
self.transformer = transformer
self.N = 5
def __getitem__(self, index):
filename = self.filelist[index]
pathA = os.path.join(self.img_path, filename + '.jpg')
image = Image.open(pathA)
targets = []
for i in range(self.N):
pathB = os.path.join(self.skt_path, '%s_%02d.png' % (filename, i+1))
target = Image.open(pathB)
targets.append(target)
if self.transformer is not None:
image, targets = self.transformer(image, targets)
targets = torch.cat(targets, 0)
return image, targets
def __len__(self):
return len(self.filelist)