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
|
#!/usr/bin/env python3
# Foreground Extraction Benchmark
# Copyright 2005 Sven Neumann <sven@gimp.org>
#
"""
This is a from-scratch implementation of the benchmark proposed in
"GrabCut": interactive foreground extraction using iterated graph
cuts published in the Proceedings of the 2004 SIGGRAPH Conference.
No guarantee is made that this benchmark produces the same results
as the cited benchmark but the goal is that it does. So if you find
any bugs or inaccuracies in this code, please let us know.
The benchmark has been adapted work with the MATTING algorithm,
which is (currently) the only
implementation of gimp_drawable_foreground_extract(). If other
implementations are being added, this benchmark should be changed
accordingly.
You will need a set of test images to run this benchmark, preferably
the original set of 50 images. Some of these images are from the
Berkeley Segmentation Dataset
http://www.cs.berkeley.edu/projects/vision/grouping/segbench/ .
See http://www.siox.org/details.html to download trimaps.
See https://web.archive.org/web/20050209123253/http://research.microsoft.com/vision/cambridge/segmentation/
and download the "Labelling - Lasso" file.
"""
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <https://www.gnu.org/licenses/>.
import os, re, struct, sys, time
import gi
gi.require_version('Gimp', '3.0')
from gi.repository import Gimp
from gi.repository import GObject
from gi.repository import GLib
from gi.repository import Gio
def benchmark (procedure, args, data):
if args.length() != 3:
error = 'Wrong parameters given'
return procedure.new_return_values(Gimp.PDBStatusType.CALLING_ERROR,
GLib.Error(error))
run_mode = args.index(0)
folder = args.index(1)
save_output = args.index(2)
folder = os.path.abspath(os.path.expanduser(folder))
if not os.path.exists(folder):
error = "Folder '" + folder + "' doesn't exist.\n"
return procedure.new_return_values(Gimp.PDBStatusType.CALLING_ERROR,
GLib.Error(error))
total_unclassified = 0
total_misclassified = 0
total_time = 0.0
images = os.path.join(folder, "images")
for name in os.listdir(images):
try:
image_display.delete()
mask_display.delete()
except NameError:
pass
image_name = os.path.join (images, name)
# Remove suffix, assuming it has three characters
name = re.sub(r'\....$', '', name)
mask_name = os.path.join(folder, "cm_bmp", name + '.png')
truth_name = os.path.join(folder, "truth", name + '.bmp')
image = Gimp.file_load(run_mode, Gio.file_new_for_path(image_name))
image_layer = image.get_active_layer()
mask = Gimp.file_load(run_mode, Gio.file_new_for_path(mask_name))
convert_grayscale(mask)
mask_layer = mask.get_active_layer()
truth = Gimp.file_load(run_mode, Gio.file_new_for_path(truth_name))
convert_grayscale(truth)
truth_layer = truth.get_active_layer()
unclassified = unclassified_pixels(mask_layer, truth_layer)
sys.stderr.write(os.path.basename (image_name))
start = time.time()
image_layer.foreground_extract(Gimp.ForegroundExtractMode.MATTING, mask_layer)
end = time.time()
sys.stderr.write(" ")
# This line was in the gimp 2 implementation, and probably isn't needed anymore.
# mask_layer.flush ()
# Ignore errors when creating image displays;
# allows us to be used without a display.
try:
image_display = Gimp.Display.new(image)
mask_display = Gimp.Display.new(mask)
Gimp.displays_flush()
time.sleep(1.0)
except:
pass
image.delete()
misclassified = misclassified_pixels (mask_layer, truth_layer)
sys.stderr.write("%d %d %.2f%% %.3fs\n" %
(unclassified, misclassified,
(misclassified * 100.0 / unclassified),
end - start))
total_unclassified += unclassified
total_misclassified += misclassified
total_time += end - start
truth.delete()
if save_output:
filename = os.path.join(folder, "output", name + '.png')
Gimp.file_save(Gimp.RunMode.NONINTERACTIVE, mask, mask_layer, Gio.file_new_for_path(filename))
mask.delete()
# for loop ends
try:
image_display.delete()
mask_display.delete()
except NameError:
pass
sys.stderr.write("Total: %d %d %.2f%% %.3fs\n" %
(total_unclassified, total_misclassified,
(total_misclassified * 100.0 / total_unclassified),
total_time))
return procedure.new_return_values(Gimp.PDBStatusType.SUCCESS, GLib.Error())
def convert_grayscale(image):
if not image.get_effective_color_profile().is_gray():
image.convert_grayscale()
def unclassified_pixels(mask, truth):
(result, mean, std_dev, median, pixels,
count, percentile) = mask.histogram(Gimp.HistogramChannel.VALUE, 2/256.0, 254/256.0)
return count
def misclassified_pixels(mask, truth):
image = truth.get_image()
copy = Gimp.Layer.new_from_drawable(mask, image)
copy.set_name("Difference")
copy.set_mode(Gimp.LayerMode.DIFFERENCE_LEGACY)
image.insert_layer(copy, None, -1)
# The assumption made here is that the output of
# foreground_extract is a strict black and white mask. The truth
# however may contain unclassified pixels. These are considered
# unknown, a strict segmentation isn't possible here.
#
# The result of using the Difference mode as done here is that
# pure black pixels in the result can be considered correct.
# White pixels are wrong. Gray values were unknown in the truth
# and thus are not counted as wrong.
flat_image = image.flatten()
(result, mean, std_dev, median, pixels,
count, percentile) = flat_image.histogram(Gimp.HistogramChannel.VALUE, 254/256.0, 1.0)
return count
PROCNAME = "python-fu-benchmark-foreground-extract"
class BenchmarkForegroundExtract(Gimp.PlugIn):
## Parameters ##
__gproperties__ = {
"run-mode": (Gimp.RunMode,
"Run mode",
"The run mode",
Gimp.RunMode.NONINTERACTIVE,
GObject.ParamFlags.READWRITE),
"image_folder": (str,
"Image Folder",
"Image Folder",
"~/segmentation/msbench/imagedata",
GObject.ParamFlags.READWRITE),
"save_output": (bool,
"Save output images",
"Save output images",
False,
GObject.ParamFlags.READWRITE)
}
## GimpPlugIn virtual methods ##
def do_query_procedures(self):
self.set_translation_domain("gimp30-python",
Gio.file_new_for_path(Gimp.locale_directory()))
return [PROCNAME]
def do_create_procedure(self, name):
procedure = None
if name == PROCNAME:
procedure = Gimp.Procedure.new(self, name,
Gimp.PDBProcType.PLUGIN,
benchmark, None)
procedure.set_documentation(
"Benchmark and regression test for Foreground Extraction",
globals()["__doc__"], # This includes the docstring, on the top of the file
name)
procedure.set_menu_label("Foreground Extraction")
procedure.set_attribution("Sven Neumann",
"Sven Neumann",
"2005")
procedure.add_menu_path("<Image>/Filters/Extensions/Benchmark")
procedure.add_argument_from_property(self, "run-mode")
procedure.add_argument_from_property(self, "image_folder")
procedure.add_argument_from_property(self, "save_output")
return procedure
Gimp.main(BenchmarkForegroundExtract.__gtype__, sys.argv)
|
