-
Notifications
You must be signed in to change notification settings - Fork 5
/
tile_im.py
257 lines (210 loc) · 9.02 KB
/
tile_im.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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
Created on Thu Aug 23 15:55:35 2018
@author: avanetten
"""
from __future__ import print_function
import os
import time
import argparse
import numpy as np
import pandas as pd
import cv2
# cv2 can't load large files, so need to import skimage too
import skimage.io
#import sys
#path_basiss = os.path.dirname(os.path.realpath(__file__))
#sys.path.append(path_basiss)
#import basiss
###############################################################################
def slice_ims(im_dir, out_dir, slice_x, slice_y,
stride_x, stride_y,
pos_columns = ['idx', 'name', 'name_full', 'xmin',
'ymin', 'slice_x',
'slice_y', 'im_x', 'im_y'],
sep0='__', sep1='_', pad=0, ext='.tif', verbose=True):
'''Slice images into patches, assume ground truth masks
are present
Adapted from basiss.py'''
if verbose:
print ("Slicing images in:", im_dir)
t0 = time.time()
count = 0
pos_list, name_list = [], []
#nims,h,w,nbands = im_arr.shape
im_roots = [z for z in os.listdir(im_dir) if z.endswith('.tif')]
#im_roots = ['90.tif']
for i,im_root in enumerate(im_roots):
im_path = os.path.join(im_dir, im_root)
if verbose:
print (i, "/", len(im_roots), "im_path:", im_path)
name = im_root.split('.')[0]
use_skimage = False
try:
# cv2 can't load large files
im = cv2.imread(im_path)
except:
# load with skimage, (reversed order of bands)
im = skimage.io.imread(im_path)#[::-1]
use_skimage = True
h, w, nbands = im.shape
print ("im.shape:", im.shape)
seen_coords = set()
#if verbose and (i % 10) == 0:
# print (i, "im_root:", im_root)
# dice it up
# after resize, iterate through image
# and bin it up appropriately
for x in range(0, w, stride_x):
for y in range(0, h, stride_y):
xmin = max(0, min(x, w-slice_x) )
ymin = max(0, min(y, h - slice_y) )
coords = (xmin, ymin)
# check if we've already seen these coords
if coords in seen_coords:
continue
else:
seen_coords.add(coords)
# check if we screwed up binning
if (slice_x <= w and (xmin + slice_x > w)) \
or (slice_y <= h and (ymin + slice_y > h)):
print ("Improperly binned image,")
return
# get satellite image cutout
im_cutout = im[ymin:ymin + slice_y,
xmin:xmin + slice_x]
##############
# skip if the whole thing is black
if np.max(im_cutout) < 1.:
continue
else:
count += 1
if verbose and (count % 50) == 0:
print ("count:", count, "x:", x, "y:", y)
###############
# set slice name
name_full = name + sep0 + str(ymin) + sep1 + str(xmin) + sep1 \
+ str(slice_y) + sep1 + str(slice_x) + sep1 + str(pad) + sep1 \
+ str(w) + sep1 + str(h) + ext
##name_full = str(i) + sep + name + sep \
#name_full = name + sep \
# + str(xmin) + sep + str(ymin) + sep \
# + str(slice_x) + sep + str(slice_y) \
# + sep + str(w) + sep + str(h) \
# + '.tif'
pos = [i, name, name_full, xmin, ymin, slice_x, slice_y, w, h]
# add to arrays
#idx_list.append(idx_full)
name_list.append(name_full)
#im_list.append(im_cutout)
#mask_list.append(mask_cutout)
pos_list.append(pos)
name_out = os.path.join(out_dir, name_full)
if not use_skimage:
cv2.imwrite(name_out, im_cutout)
else:
# if we read in with skimage, need to reverse colors
cv2.imwrite(name_out, cv2.cvtColor(im_cutout, cv2.COLOR_RGB2BGR))
# create position datataframe
df_pos = pd.DataFrame(pos_list, columns=pos_columns)
df_pos.index = np.arange(len(df_pos))
if verbose:
print (" len df;", len(df_pos))
print (" Time to slice arrays:", time.time() - t0, "seconds")
return df_pos
###############################################################################
def main():
# construct the argument parse and parse the arguments
parser = argparse.ArgumentParser()
parser.add_argument('--im_dir', type=str, default='/ims/to/tile/',
help="images location")
parser.add_argument('--out_dir', type=str, default='/output/folder/for/tiles/',
help="output_images location")
parser.add_argument('--slice_x', type=int, default=544)
parser.add_argument('--slice_y', type=int, default=544)
parser.add_argument('--stride_x', type=int, default=108)
parser.add_argument('--stride_y', type=int, default=108)
args = parser.parse_args()
if not os.path.exists(args.out_dir):
os.mkdir(args.out_dir)
df_pos = slice_ims(args.im_dir, args.out_dir, args.slice_x, args.slice_y,
args.stride_x, args.stride_y,
pos_columns = ['idx', 'name', 'name_full', 'xmin',
'ymin', 'slice_x',
'slice_y', 'im_x', 'im_y'],
verbose=True)
path_tile_df_csv = os.path.join(os.path.dirname(args.out_dir), os.path.basename(args.out_dir) + '_df.csv')
# save to file
df_pos.to_csv(path_tile_df_csv)
print ("df saved to file:", path_tile_df_csv)
# # use config file
# # use config file?
# from config import Config
# import json
# parser = argparse.ArgumentParser()
# parser.add_argument('config_path')
# args = parser.parse_args()
# # get config
# with open(args.config_path, 'r') as f:
# cfg = json.load(f)
# config = Config(**cfg)
#
# # get input dir
# path_images_8bit = os.path.join(config.path_data_root, config.test_data_refined_dir)
#
# # make output dirs
# # first, results dir
# res_dir = os.path.join(config.path_results_root, config.test_results_dir)
# os.makedirs(res_dir, exist_ok=True)
# path_tile_df_csv = os.path.join(config.path_results_root, config.test_results_dir, config.tile_df_csv)
# path_tile_df_csv2 = os.path.join(config.path_data_root, os.path.dirname(config.test_sliced_dir), config.tile_df_csv)
#
# # path for sliced data
# path_sliced = os.path.join(config.path_data_root, config.test_sliced_dir)
#
# #if not os.path.exists(config.results_dir):
# # os.mkdir(config.results_dir)
# #if not os.path.exists(config.path_sliced):
# # os.mkdir(config.path_sliced)
#
# # only run if nonzer tile and sliced_dir
# if (len(config.test_sliced_dir) > 0) and (config.slice_x > 0):
# os.makedirs(path_sliced, exist_ok=True)
#
#
# df_pos = slice_ims(path_images_8bit, path_sliced,
# config.slice_x, config.slice_y,
# config.stride_x, config.stride_y,
# pos_columns = ['idx', 'name', 'name_full', 'xmin',
# 'ymin', 'slice_x',
# 'slice_y', 'im_x', 'im_y'],
# verbose=True)
# # save to file
# df_pos.to_csv(path_tile_df_csv)
# print ("df saved to file:", path_tile_df_csv)
# # also csv save to data dir
# df_pos.to_csv(path_tile_df_csv2)
# # iterate through im_dir and gather files
# im_arr = []
# name_arr = []
# mask_arr = []
# im_roots = [z for z in os.listdir(args.im_dir) of z.endswith('.tif')]
# for i,im_root in enumerate(im_roots):
# im_file_name = os.path.join(im_dir, im_root)
# im = cv2.imread(im_file_name, 1)
# name_arr.append(im_file_name)
# im_arr.append(im)
#
# # slice
# df_pos, name_out_arr, im_out_arr, mask_out_arr = \
# basiss.slice_ims(im_arr, mask_arr, names_arr,
# args.slice_x, args.slice_y,
# args.stride_x, args.stride_y,
# pos_columns = ['idx', 'name', 'xmin',
# 'ymin', 'slice_x',
# 'slice_y', 'im_x', 'im_y'],
# verbose=True)
###############################################################################
if __name__ == '__main__':
main()