-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathdataloader.py
160 lines (116 loc) · 6.37 KB
/
dataloader.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
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
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
import os
import torch
import random
import numpy as np
from tqdm import tqdm
# from osgeo import gdal
import rasterio
from os.path import dirname as up
from torch.utils.data import Dataset
import torchvision.transforms.functional as F
import pytorch_lightning as pl
import pandas as pd
random.seed(0)
np.random.seed(0)
torch.manual_seed(0)
# NAIP data
bands_std = np.array([53.4146, 53.4118, 53.4115, 53.4158]).astype('float32')
bands_mean = np.array([98.8948, 98.8902, 98.8835, 98.8795]).astype('float32')
# # Setting dataset folder and its weights
# # ********************************************************************************************
dataset_dir = os.path.join(up(__file__), 'datasets') #, '256_images_lesstrain'
# # # Pixel-Level class distribution for each dataset
# """
# {'Image_allyear_merged_512': array([0.49019898, 0.44645175, 0.02657669, 0.03677258]),
# 'Image_after_2010_merged_512': array([0.45526231, 0.43276168, 0.05267328, 0.05930273]),
# 'Image_after_2010_merged_256': array([0.32490837, 0.41855632, 0.17178225, 0.08475305]),
# 'Image_allyear_merged_256': array([0.44056167, 0.41559786, 0.09143975, 0.05240071]),
# 'Image_allyear_merged_1024': array([0.50426896, 0.4565353 , 0.01033589, 0.02885985]),
# 'Image_after_2010_merged_1024': array([0.49582348, 0.43486081, 0.02433212, 0.04498359]),
# 'Image_after_2010_VA_512': array([0.47100772, 0.44772889, 0.01990966, 0.06135373]),
# 'Image_after_2010_VA_256': array([0.37283283, 0.48029399, 0.04961892, 0.09725425]),
# 'Image_allyear_VA_512': array([0.49843776, 0.45395527, 0.01021636, 0.03739061]),
# 'Image_allyear_VA_256': array([0.47332474, 0.44650446, 0.02387324, 0.05629757])}
# """
# if data_name == 'Image_after_2010_merged_512':
# class_distr = torch.Tensor([0.45526231, 0.43276168, 0.05267328, 0.05930273]) # 4 classes
# elif data_name == 'Image_after_2010_merged_256':
# class_distr = torch.Tensor([0.32490837, 0.41855632, 0.17178225, 0.08475305])
# elif data_name == 'Image_after_2010_merged_1024':
# class_distr = torch.Tensor([0.49582348, 0.43486081, 0.02433212, 0.04498359])
# elif data_name == 'Image_allyear_merged_256':
# class_distr = torch.Tensor([0.44056167, 0.41559786, 0.09143975, 0.05240071])
# elif data_name == 'Image_allyear_merged_512':
# class_distr = torch.Tensor([0.49019898, 0.44645175, 0.02657669, 0.03677258])
# elif data_name == 'Image_allyear_merged_1024':
# class_distr = torch.Tensor([0.50426896, 0.4565353 , 0.01033589, 0.02885985])
# elif data_name == 'Image_after_2010_VA_512':
# class_distr = torch.Tensor([0.47100772, 0.44772889, 0.01990966, 0.06135373])
# elif data_name == 'Image_after_2010_VA_256':
# class_distr = torch.Tensor([0.37283283, 0.48029399, 0.04961892, 0.09725425])
# elif data_name == 'Image_allyear_VA_512':
# class_distr = torch.Tensor([0.49843776, 0.45395527, 0.01021636, 0.03739061])
# elif data_name == 'Image_allyear_VA_256':
# class_distr = torch.Tensor([0.47332474, 0.44650446, 0.02387324, 0.05629757])
# else:
# raise
###############################################################
# Pixel-level Semantic Segmentation Data Loader #
###############################################################
class ShorelineArmoring(Dataset): # Extend PyTorch's Dataset class
def __init__(self, mode = 'train', transform=None, standardization=None, data_name = "Image_after_2010_merged_256", path = dataset_dir):
if os.path.isdir(os.path.join(path, data_name)):
data_path = os.path.join(path, data_name)
print(data_path)
else:
raise
if mode=='train':
self.ROIs = np.array([name for name in os.listdir(os.path.join(data_path, 'train')) if os.path.isfile(os.path.join(data_path, 'train', name))])
elif mode=='test':
self.ROIs = np.array([name for name in os.listdir(os.path.join(data_path, 'test')) if os.path.isfile(os.path.join(data_path, 'test', name))])
elif mode=='val':
self.ROIs = np.array([name for name in os.listdir(os.path.join(data_path, 'val')) if os.path.isfile(os.path.join(data_path, 'val', name))])
else:
raise
self.X = [] # Loaded Images
self.y = [] # Loaded Output masks
for roi in tqdm(self.ROIs, desc = 'Load '+mode+' set to memory'):
# Construct file and folder name from roi
roi_file = os.path.join(data_path, mode, roi)
roi_file_mask = os.path.join(data_path, 'masks', roi)
# Load Classsification Mask
ds = rasterio.open(roi_file_mask)
temp = np.copy(ds.read().astype(np.int64))
# Categories from 1 to 0
temp = np.copy(temp - 1)
ds=None
self.y.append(temp)
# Load Patch
ds = rasterio.open(roi_file)
temp = np.copy(ds.read())
ds=None
self.X.append(temp)
self.impute_nan = np.tile(bands_mean, (temp.shape[1],temp.shape[2],1))
self.mode = mode
self.transform = transform
self.standardization = standardization
self.length = len(self.y)
def __len__(self):
return self.length
def getnames(self):
return self.ROIs
def __getitem__(self, index):
img = self.X[index]
target = self.y[index]
img = np.moveaxis(img, [0, 1, 2], [2, 0, 1]).astype('float32') # CHW to HWC
nan_mask = np.isnan(img)
img[nan_mask] = self.impute_nan[nan_mask]
if self.transform is not None:
target = np.moveaxis(target, [0, 1, 2], [2, 0, 1]) # CHW to HWC
stack = np.concatenate([img, target], axis=-1).astype('float32') # In order to rotate-transform both mask and image
stack = self.transform(stack)
img = stack[:-1,:,:]
target = stack[-1,:,:].long() # Recast target values back to int64 or torch long dtype
if self.standardization is not None:
img = self.standardization(img)
return {'image': img, 'mask': target}