-
Notifications
You must be signed in to change notification settings - Fork 1
/
crop_resize.py
289 lines (219 loc) · 12.1 KB
/
crop_resize.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
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
import os
import cv2
from matplotlib import pyplot as plt
import numpy as np
import nibabel as nib
from PIL import Image, ImageDraw
def image_resize(load_image_path, save_image_path, resized_size, is_display):
img = Image.open(load_image_path)
# resized_img = img.resize(resized_size, Image.LANCZOS)
img.save(save_image_path,"PNG")
# if is_display:
# resized_img.show()
def image_crop(load_image_path, save_image_path, crop_area_path, cropped_size, is_display):
img = Image.open(load_image_path)
img_arr = np.asarray(img)
old_row, old_col = img_arr.shape
new_row, new_col = cropped_size
cropped_img_arr = img_arr[ new_row // 2 : old_row - (new_row // 2),
new_col // 2 : old_col - (new_col // 2) ]
cropped_img = Image.fromarray(np.uint8(cropped_img_arr))
# cropped_img.save(save_image_path, format='png')
# draw & save crop area
before_image = Image.open(load_image_path).convert('RGB')
draw_filter = ImageDraw.Draw(before_image)
draw_filter.rectangle((new_row//2, new_row//2, old_row-(new_row//2), old_row-(new_row//2)), outline='red', width = 3)
# before_image.show()
before_image.save(crop_area_path, format='png')
if is_display:
cropped_img.show()
def make_enface(load_nifti_path, load_mask_path, save_enface_path):
nii_img_og = nib.load(load_nifti_path)
nii_arr = np.asarray(nii_img_og.dataobj)
mask_arr = np.asarray(nib.load(load_mask_path).dataobj)
srl_arr = np.zeros(np.shape(nii_arr))
srl_arr[(0<mask_arr) & (mask_arr<5)] = nii_arr[(0<mask_arr) & (mask_arr<5)]
srl_pruned = np.zeros(np.shape(srl_arr))
lower = np.percentile(srl_arr[np.where(mask_arr > 0)], 0)
upper = np.percentile(srl_arr[np.where(mask_arr > 0)], 99.97)
srl_pruned[(lower <= srl_arr) & (srl_arr <= upper)] = srl_arr[(lower <= srl_arr) & (srl_arr <= upper)]
m = np.min(srl_pruned)
M = np.max(srl_pruned)
srl_pruned = np.uint8((srl_pruned - m) / (M - m)*255)
nii_enface = np.max(srl_pruned, axis=1)
nii_enface = np.rot90(nii_enface)
nii_img = Image.fromarray(np.uint8(nii_enface))
nii_img.save(save_enface_path, format='png')
def make_bscan(load_nifti_path, save_bscan_path, slice_num, is_superficial=None):
nii_img = nib.load(load_nifti_path)
nii_arr = nii_img.dataobj
nii_arr = np.transpose(nii_arr, (0,2,1))
nii_new = np.zeros(np.shape(nii_arr))
for idx in range(np.shape(nii_arr)[-1]):
nii_new[:,:,idx] = np.rot90(nii_arr[:,:,idx], k=3)
if slice_num is None:
slice_num = np.shape(nii_arr)[0]//2
nii_bscan = nii_new[slice_num,:,:]
if is_superficial:
nii_bscan_srl = np.zeros(np.shape(nii_bscan))
nii_bscan_srl[(nii_bscan>0) & (nii_bscan<5)] = nii_bscan[(nii_bscan>0) & (nii_bscan<5)]
nii_bscan = nii_bscan_srl
m = np.min(nii_bscan)
M = np.max(nii_bscan)
nii_bscan = np.uint8((nii_bscan-m)/(M-m)*255)
nii_bscan = np.rot90(nii_bscan, k=3)
nii_bscan_img = Image.fromarray(nii_bscan)
nii_bscan_img.save(save_bscan_path, format='png')
def get_FOV33_info(nii_size):
scale = 19*3
fov33_size = (3, 2, 3) # fov66_size = (6, 6, 2)
resolution = [(f*scale) for f in fov33_size]
spacing = [(r/v) for r,v in zip(resolution, nii_size)]
origin = [-r/2 for r in resolution]
new_info = {'pixel' : nii_size,
'origin' : origin,
'spacing': spacing}
return new_info
def save_nifti(nifti_data, before_info, after_info, save_path):
w_pix, h_pix, d_pix = after_info['pixel']
w_spacing, h_spacing, d_spacing = after_info['spacing']
w_origin, h_origin, d_origin = after_info['origin']
nifti_info_new = before_info
nifti_info_new['dim'] = [3, w_pix, h_pix, d_pix, 1, 1, 1, 1]
nifti_info_new['xyzt_units'] = 0
nifti_info_new['pixdim'] = [1., w_spacing, h_spacing, d_spacing, 1., 1., 1., 1.]
nifti_info_new['qoffset_x'] = -w_origin
nifti_info_new['qoffset_y'] = -h_origin
nifti_info_new['qoffset_z'] = d_origin
nifti_info_new['srow_x'] = [-w_spacing, 0, 0, -w_origin]
nifti_info_new['srow_y'] = [0, -h_spacing, 0, -h_origin]
nifti_info_new['srow_z'] = [0, 0, d_spacing, d_origin]
nifti_new_img = nib.Nifti1Image(np.uint8(nifti_data), nifti_info_new.get_best_affine())
nib.save(nifti_new_img, save_path, )
print(f'nifti saved at {save_path}')
def extract_patch(load_image_path):
from sklearn.feature_extraction import image
img = Image.open(load_image_path)
img_arr = np.asarray(img)[:,:,0]
img_size = img_arr.shape
patch_num = 4
row_size, col_size = (int(img_size[0]//np.sqrt(patch_num)), int(img_size[1]//np.sqrt(patch_num)))
img_1 = img_arr[0:row_size, 0:col_size]
img_2 = img_arr[row_size: , 0:col_size]
img_3 = img_arr[0:row_size, col_size:]
img_4 = img_arr[row_size: , col_size:]
pil_img_1 = Image.fromarray(np.uint8(img_1))
pil_img_2 = Image.fromarray(np.uint8(img_2))
pil_img_3 = Image.fromarray(np.uint8(img_3))
pil_img_4 = Image.fromarray(np.uint8(img_4))
pil_img_1_resized = pil_img_1.resize((100, 100), Image.NEAREST)
pil_img_2_resized = pil_img_2.resize((100, 100), Image.NEAREST)
pil_img_3_resized = pil_img_3.resize((100, 100), Image.NEAREST)
pil_img_4_resized = pil_img_4.resize((100, 100), Image.NEAREST)
print()
def main():
# subjects = [ i for i in range(10001,10501) ]
subjects = [10132, 10334]#, 10334]
subjects.extend([i for i in range(10001, 10501, 50)])
subjects = [i for i in range(10001, 10501)]
before_registered_dir = {'mask':'/data/Nifti/In/Before_Transformed/VolMask',
'octa':'/data/Nifti/In/Before_Transformed/OCTA',
'oct' :'/data/Nifti/In/Before_Transformed/OCT'}
after_registered_dir = {'mask':'/data/Nifti/In/Transformed/VolMask_reg',
'octa':'/data/Nifti/In/Transformed/OCTA',
'oct' :'/data/Nifti/In/Transformed/OCT'} # 10001_mask_translate_rigid.nii.gz
after_crop_resize_dir = {'octa' : '/data/Nifti/In/Transformed/OCTA_Crop_Resize',
'mask' : '/data/Nifti/In/Transformed/Mask_Crop_Resize'}
final_dir = {'octa' : '/data/Nifti/In/Transformed/OCTA_SRL_256',
'mask' : '/data/Nifti/In/Transformed/Mask_SRL_256'}
bscan_dir = {'mask':'/data/Nifti/In/Transformed/Bscan/VolMask',
'octa':'/data/Nifti/In/Transformed/Bscan/OCTA',
'oct' :'/data/Nifti/In/Transformed/Bscan/OCT'} # 10001_mask_translate_rigid.nii.gz
rename_dir = '/data/Nifti/In/Transformed/OCTA_SRL_256_V2'
image_dir ='/data/Nifti/In/Transformed/OCTA_SRL_EnFace'
for subject in subjects:
load_img_path = os.path.join(image_dir, f'{subject}.png')
# extract_patch(load_image_path = load_img_path)
image_resize(load_image_path = os.path.join('/data/dataset/OCTA-500/OG', f'{subject}.bmp'),
save_image_path = os.path.join('/data/dataset/OCTA-500/OG', f'{subject}.png'),
resized_size = (192, 192),
is_display = False)
# load_nii_path = os.path.join(final_dir['octa'], f'{subject}.nii.gz')
# mask_nii_path = os.path.join(final_dir['mask'], f'{subject}.nii.gz')
# make_enface(load_nifti_path = load_nii_path,
# load_mask_path = mask_nii_path,
# save_enface_path = f'/data/Nifti/In/Transformed/EnFace/{subject}_final.png',
# subject = subject)
# ### Bscan - for Z Axis C
# load_dir = '/data/Nifti/In/Transformed/Mask_SRL_256'
# load_path = os.path.join(load_dir, f'{subject}.nii.gz')
# bscan_path = os.path.join(bscan_dir['mask'], f'{subject}_mask_reg_resize_zcrop.png')
# make_bscan(load_nifti_path = load_path,
# save_bscan_path = bscan_path,
# slice_num = 192//2)
# ### Bscan - for Crop & Resize Result from FOV 3x3
# load_dir = '/data/Nifti/In/Transformed/FOV_33/Mask_192'
# resized_path = os.path.join(load_dir, f'{subject}_mask_fov33_resized.nii.gz')
# make_bscan(load_nifti_path = resized_path,
# save_bscan_path = os.path.join(bscan_dir['mask'], f'{subject}_reg_resized.png'),
# slice_num = 192//2)
# ### Bscan - for Crop & Resize Result from FOV 6x6
# if subject<10301:
# load_dir = '/data/Nifti/In/Transformed/FOV_66/Mask_192'
# cropped_path = os.path.join(load_dir, f'{subject}_mask_fov66_cropped.nii.gz')
# resized_path = os.path.join(load_dir, f'{subject}_mask_fov66_resized.nii.gz')
# else:
# load_dir = '/data/Nifti/In/Transformed/FOV_33/Mask_192'
# cropped_path = os.path.join(load_dir, f'{subject}_mask_fov33_cropped.nii.gz')
# resized_path = os.path.join(load_dir, f'{subject}_mask_fov33_resized.nii.gz')
# make_bscan(load_nifti_path = cropped_path,
# save_bscan_path = os.path.join(bscan_dir['mask'], f'{subject}_reg_crop.png'),
# slice_num = 200//2)
# make_bscan(load_nifti_path = resized_path,
# save_bscan_path = os.path.join(bscan_dir['mask'], f'{subject}_reg_crop_resize.png'),
# slice_num = 192//2)
# make_bscan(load_nifti_path = resized_path,
# save_bscan_path = os.path.join(bscan_dir['mask'], f'{subject}_reg_crop_resize_srl.png'),
# slice_num = 192//2,
# is_superficial = True)
'''
# These parts are for making bscan - for before & after registration
slice_num = 400//2 if subject<10301 else 304//2
make_bscan(load_nifti_path = os.path.join(before_registered_dir['mask'],f'{subject}.nii.gz'),
save_bscan_path = os.path.join(bscan_dir['mask'],f'{subject}_reg_before.png'),
slice_num = slice_num)
make_bscan(load_nifti_path = os.path.join(after_registered_dir['mask'],f'{subject}_mask_translate_rigid.nii.gz'),
save_bscan_path = os.path.join(bscan_dir['mask'],f'{subject}_reg_after.png'),
slice_num = slice_num)
'''
'''
#### These parts are for the Crop & Resizing 2D ground truth projected Image ####
if subject < 10301:
before_crop_dir = '/data/dataset/OCTA-500/OCTA_6M/Projection Maps/OCTA(ILM_OPL)'
after_crop_dir = '/data/dataset/OG/OCTA_Enface'
crop_area_dir = '/data/dataset/OG/OCTA_Enface_with_crop_area'
os.makedirs(crop_area_dir, exist_ok=True)
#### Make En Face Image ####
make_enface(load_nifti_path=f'/data/Nifti/In/FOV_66/OCTA_srl/{subject}.nii.gz',
save_enface_path=f'/home/dhl/Project/OCTPreprocessing/enface/{subject}.png')
#### Crop 2D Image ####
image_crop(load_image_path = os.path.join(before_crop_dir, f'{subject}.bmp'),
save_image_path = os.path.join(after_crop_dir, f'{subject}.png'),
crop_area_path = os.path.join(crop_area_dir,f'{subject}.png'),
cropped_size = (200, 200),
is_display = False)
#### Resize 2D Image ####
image_resize(load_image_path = os.path.join(after_crop_dir, f'{subject}.png'),
save_image_path = os.path.join(after_crop_dir, f'{subject}.png'),
resized_size = (192, 192),
is_display = False)
else:
before_resize_dir = '/data/dataset/OCTA-500/OCTA_3M/Projection Maps/OCTA(ILM_OPL)'
after_resize_dir = '/data/dataset/OCTA_Enface'
image_resize(load_image_path = os.path.join(before_resize_dir, f'{subject}.bmp'),
save_image_path = os.path.join(after_resize_dir, f'{subject}.png'),
resized_size = (192, 192),
is_display = False)
'''
print(f'{subject} has been saved.')
main()