-
Notifications
You must be signed in to change notification settings - Fork 3
/
run_dataset_generator_for_two_view_scanner.py
317 lines (259 loc) · 13.2 KB
/
run_dataset_generator_for_two_view_scanner.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
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
# -----------------------------------------------------------------------------
"""run_debisim_dataset_generator.py: Run the generator for creating randomized
CT datasets using the DEBISim pipeline.
"""
__author__ = "Ankit Manerikar"
__copyright__ = "Copyright (C) 2021, Robot Vision Lab"
__date__ = "12th May, 2023"
__credits__ = ["Ankit Manerikar", "Fangda Li"]
__license__ = "Public Domain"
__version__ = "2.1.0"
__maintainer__= ["Ankit Manerikar", "Fangda Li"]
__email__ = ["amanerik@purdue.edu", "li1208@purdue.edu"]
__status__ = "Prototype"
# -----------------------------------------------------------------------------
import warnings
warnings.filterwarnings('ignore')
import argparse
import importlib.util as config_loader
import warnings
warnings.filterwarnings('ignore')
from src.debisim_pipeline import *
from lib.decomposer.cdm_decomposer import *
from lib.misc.benchmark import *
import torchvision.transforms as tvt
img_suffixes = dict(
cdm=['c', 'pe', 'z'],
sirz=['ze', 'rho']
)
parser = argparse.ArgumentParser(
description='Dataset Generator for DEBISim: \n'
'-----------\n'
'The script generates a simulated Two-view X-ray'
'dataset of randomized baggage configurations. '
'The simulation parameters are specified using '
'a config.py file - these include setting up '
'the scanner + X-ray source/detector, '
'the types of objects of objects to be spawned '
'in the bag as well as the DE decomposition '
'+ reconstruction parameters.'
'Examples of config.py files are provided in '
'configs/ directory for different scanners '
'and scanner geometries')
parser.add_argument('--config',
default=os.path.join(CONFIG_DIR,
'config_default_two_view_conebeam_dect.py'),
help='config file location',
dest='config'
)
parser.add_argument('--sim_dir',
default=os.path.join(RESULTS_DIR,
'example_default_two_view_conebeam_dect/'),
help='simulation directory for saving output'
)
parser.add_argument('--num_bags',
default=10,
help='number of bags to simulate',
type=int
)
args = parser.parse_args()
def run_two_view_xray_dataset_generator(num_bags,
sim_dir,
scanner,
xray_src_mdl,
bag_creator_args,
decomposer='cdm',
save_sino=False,
basis_fn=None,
decomposer_args=None,
images_to_save=None,
slicewise=True,
compress_data=False,
fwd_mdl_args=None,
rotate_bag=False
):
"""
---------------------------------------------------------------------------
Function to generate a CT baggage Dataset using the DebiSim pipeline.
:param num_bags: number of baggage instances to generate
:param sim_dir: simulation directory
:param scanner: scanner model defined from ScannerTemplate()
:param xray_src_mdl: Xray source specifications - these are defined
as a dictionary with the following
specifications:
- 'num_spectra' - No of X-ray sources/spectra
- 'kVp' - peak kV voltage for the X-ray source(s)
- 'spectra' - file paths for the each of the
X-ray spectra. The spectrum files
must contain a N x 2 array with the
keV values in the first column
and normalized photon distribution
in the 2nd column. See
/include/spectra/ for reference.
- 'dosage' - dosage count for each of the sources
:param bag_creator_args: The baggage creation arguments - this dictionary
contains the input arguments to the method
self.create_random_object_list() for
BaggageCreator3D() or BaggageCreator2d()
:param decomposer: The object(s) corresponding the DE Decomposer,
for e.g., CDMDecomposer().
:param decomposer_args: Input arguments for the specified decomposer(s)
:param save_sino: If simulated sinograms are to be saved
:param basis_fn: Energy basis functions for DECT processing
:param images_to_save: ground truth images to be saved:
{gt | lac_1 | lac_2 | compton | pe | zeff}
:param slicewise: whether to produce H x W X D 3D volumetric bag
or H x W 2D baggage cross-sections in D batches
:param compress_data: whether to compress saved FITS data
:param fws_model_args arguments for X-ray forward model
:return:
---------------------------------------------------------------------------
"""
if images_to_save is None:
images_to_save = ['lac_1', 'pe', 'c', 'gt']
if fwd_mdl_args is None:
fwd_mdl_args = dict(
add_poisson_noise=True,
add_system_noise=True,
system_gain=0.0025
)
# Organize processing with Benchmark
bench = Benchmark(save_log=False,
save_remark=False)
bench.set_remark(
'Creating Randomized DEBISIM Dataset %s with %i bags ....'%(sim_dir,
len(num_bags)))
sim_bag_dir = os.path.join(sim_dir, 'simulation_%03d/')
bench.set_test_cases([sim_bag_dir % i for i in num_bags])
bench.set_output_dir([sim_bag_dir % i for i in num_bags])
# Process slice by slice
def preprocess(test_case):
# ---------------------------------------------------------------------
# Setup the simulator
# ---------------------------------------------------------------------
pre = dict(test_case=test_case)
return pre
# ---------------------------------------------------------------------
def run(pre):
# ---------------------------------------------------------------------
# Run the simulator
# ---------------------------------------------------------------------
bag_dir = pre['test_case']
# Initialized the simulator
simulator = DEBISimPipeline(
sim_path=bag_dir,
scanner_model=scanner,
xray_source_model=xray_src_mdl,
compress_data=compress_data
)
simulator.logger.info("="*80)
simulator.logger.info(" "*20+"DEBISIM PIPELINE STARTS"+" "*20)
simulator.logger.info("="*80)
mu = simulator.mu.material('water')
simulator.logger.info('\n'+"-" * 50+"BAG_GENERATOR"+"-" * 50+'\n')
simulator.create_random_simulation_instance(bag_creator_args,
save_images=images_to_save,
slicewise=slicewise)
# TODO random_rotation of bag
table_height = 448
bh, bw, bd = simulator.gt_image_3d.shape
center_of_rotation = (np.random.randint(bh//2-50, bh//2+50),
np.random.randint(bd//2-50, bd//2+50))
if rotate_bag:
bag_chunk = simulator.gt_image_3d[:table_height, :, :].clone()
bag_chunk = tvt.RandomRotation((-30,30),
center=center_of_rotation
)(bag_chunk.permute(1, 0, 2))
bag_chunk = bag_chunk.permute(1,0,2)
simulator.gt_image_3d[:table_height,:, :] = bag_chunk.clone()
# simulator.keV_range = range(50, 75)
simulator.logger.info('\n'+"-" * 50+"FORWARD_MODEL"+"-" * 50+'\n')
simulator.run_fwd_model(**fwd_mdl_args)
if simulator.xray_source_model['num_spectra'] == 1:
two_view_image = simulator.data.copy()
top_view_image, side_view_image = two_view_image[:,:,1], \
two_view_image[:,:,2]
save_fits_data(os.path.join(simulator.f_loc['image_dir'],
'bag_top_view.fits.gz'),
top_view_image)
save_fits_data(os.path.join(simulator.f_loc['image_dir'],
'bag_side_view.fits.gz'),
side_view_image)
quick_imshow(2, 1,
[top_view_image, side_view_image],
colorbar=False,
colormap='gist_yarg',
titles=['Top', 'Side'])
plt.savefig(os.path.join(simulator.f_loc['image_dir'],
'baggage_scans.png'))
plt.close()
elif simulator.xray_source_model['num_spectra'] == 2:
two_view_image_1 = simulator.data1.copy()
top_view_image_1, side_view_image_1 = two_view_image_1[:,:,1], \
two_view_image_1[:,:,2]
save_fits_data(os.path.join(simulator.f_loc['image_dir'],
'bag_top_view_spec1.fits.gz'),
top_view_image_1)
save_fits_data(os.path.join(simulator.f_loc['image_dir'],
'bag_side_view_spec1.fits.gz'),
side_view_image_1)
quick_imshow(2, 1,
[top_view_image_1, side_view_image_1],
colorbar=False,
colormap='gist_yarg',
titles=['Top (Spectrum 1)', 'Side (Spectrum 1)'])
plt.savefig(os.path.join(simulator.f_loc['image_dir'],
'baggage_scans_spec1.png'))
plt.close()
two_view_image_2 = simulator.data2.copy()
top_view_image_2, side_view_image_2 = two_view_image_2[:,:,1], \
two_view_image_2[:,:,2]
save_fits_data(os.path.join(simulator.f_loc['image_dir'],
'bag_top_view_spec2.fits.gz'),
top_view_image_2)
save_fits_data(os.path.join(simulator.f_loc['image_dir'],
'bag_side_view_spec2.fits.gz'),
side_view_image_2)
quick_imshow(2, 1,
[top_view_image_2, side_view_image_2],
colorbar=False,
colormap='gist_yarg',
titles=['Top (Spectrum 2)', 'Side (Spectrum 2)'])
plt.savefig(os.path.join(simulator.f_loc['image_dir'],
'baggage_scans_spec2.png'))
plt.close()
f_loc = simulator.f_loc.copy()
torch.cuda.empty_cache()
if decomposer=='none':
pass
else:
simulator.logger.info('\n'+'-'*50+"DECOMPOSER"+'-'*50+'\n')
simulator.run_decomposer(type=decomposer,
decomposer_args=decomposer_args,
basis_fn=basis_fn,
save_sino=save_sino)
res = dict()
if (simulator.logger.hasHandlers()):
simulator.logger.handlers.clear()
simulator.logger.propagate = False
simulator.scanner.logger.propagate = False
del simulator.logger
del simulator
return res
# -------------------------------------------------------------------------
def postprocess(res, outdir):
if res is not None:
for k, v in res.items():
del v
del res
return None
bench.set_handles(preprocess, run, postprocess, None)
bench.start()
# ----------------------------------------------------------------------------
spec = config_loader.spec_from_file_location("config.params",
args.config)
config = config_loader.module_from_spec(spec)
spec.loader.exec_module(config)
config.params['sim_dir'] = args.sim_dir
config.params['num_bags'] = range(1, args.num_bags+1)
run_two_view_xray_dataset_generator(**config.params)