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
|
'''
Runs unit tests for continuum processing functions.
To run the unit tests, type the following from the system command line:
# python -m spectral.tests.continuum
'''
from __future__ import absolute_import, division, print_function, unicode_literals
import numpy as np
from numpy.testing import assert_allclose
import spectral as spy
from spectral.algorithms.spymath import matrix_sqrt
from spectral.algorithms.continuum import spectral_continuum, remove_continuum, continuum_points
from spectral.tests.spytest import SpyTest
class ContinuumTest(SpyTest):
def setup(self):
self.image = spy.open_image('92AV3C.lan')
self.bands = np.sort(
spy.aviris.read_aviris_bands('92AV3C.spc').centers)
class FindContinuumTest(ContinuumTest):
'''Tests spectral_continuum.'''
def test_few_simple_cases(self):
spectrum = np.array([1., 2., 2.5, 1.6, 0.75, 1.5, 2.2, 2.9, 1.8])
bands = np.array([1., 2., 3., 4., 5., 6., 7., 8., 9.])
expected = np.array([1., 2., 2.5, 2.58, 2.66, 2.74, 2.82, 2.9, 1.8])
assert_allclose(expected, spectral_continuum(spectrum, bands))
spectrum = np.array([0.6, 1., 2.45, 3.1, 3.25, 4.15,
4.35, 4.1, 3.1, 2.7, 2., 2.85, 3.75, 3., 2., 0.9])
bands = np.array([0.3, 1., 1.8, 3., 4.5, 5.2, 6.45,
7., 7.1, 8., 8.1, 9., 9.3, 10.2, 10.5, 10.6])
expected = np.array([0.6, 1.46333333, 2.45, 3.1, 3.81590909, 4.15, 4.35, 4.23421053,
4.21315789, 4.02368421, 4.00263158, 3.81315789, 3.75, 3., 2., 0.9])
assert_allclose(expected, spectral_continuum(spectrum, bands))
spectrum = np.array([0.5, 1.1, 1.5, 2.4, 1.9, 1.0])
bands = np.array([0.5, 1.0, 1.7, 2.0, 3.0, 3.5])
expected = np.array([0.5, 1.13333333, 2.02, 2.4, 1.9, 1.])
assert_allclose(expected, spectral_continuum(spectrum, bands))
spectrum = np.array([0.5, 1.1, 1.8, 2.0, 1.1, 0.9, 0.4])
bands = np.array([0.5, 0.9, 1.6, 2.0, 2.1, 2.8, 3.0])
expected = np.array([0.5, 1.1, 1.8, 2., 1.8625, 0.9, 0.4])
assert_allclose(expected, spectral_continuum(spectrum, bands))
def test_simple_segmented(self):
# A case without local maximum inside concave regions.
# Same as convex hull.
spectrum = np.array([0.6, 1., 2.45, 3.1, 3.25, 4.15,
4.35, 4.1, 3.1, 2.7, 2., 2.85, 3.75, 3., 2., 0.9])
bands = np.array([0.3, 1., 1.8, 3., 4.5, 5.2, 6.45,
7., 7.1, 8., 8.1, 9., 9.3, 10.2, 10.5, 10.6])
expected = np.array([0.6, 1.46333333, 2.45, 3.1, 3.81590909, 4.15, 4.35,
4.23421053, 4.21315789, 4.02368421, 4.00263158, 3.81315789, 3.75, 3.,
2., 0.9, ])
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# A case with single local maxima that gets filtered out, because it
# does not satisfy quasi-convexity.
spectrum = np.array([0.6, 1., 2.45, 3.1, 3.25, 4.15,
4.35, 4.1, 3.1, 3.7, 2., 2.85, 3.75, 3., 2., 0.9])
bands = np.array([0.3, 1., 1.8, 3., 4.5, 5.2, 6.45,
7., 7.1, 8., 8.1, 9., 9.3, 10.2, 10.5, 10.6])
expected = np.array([0.6, 1.46333333, 2.45, 3.1, 3.81590909,
4.15, 4.35, 4.23421053, 4.21315789, 4.02368421,
4.00263158, 3.81315789, 3.75, 3., 2.,
0.9])
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# Reversed case.
spectrum = spectrum[::-1]
bands = np.cumsum(np.concatenate(
(np.array([0.3]), (bands[1:] - bands[:-1])[::-1])))
expected = expected[::-1]
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# A case with single valid local maxima inside concave region,
# and one invalid.
spectrum = np.array([0.60, 1.00, 2.45, 3.10, 3.25, 4.15, 4.35, 4.10,
3.10, 3.80, 3.50, 3.60, 2.00, 2.85, 3.75, 3.00, 2.00, 0.90])
bands = np.array([0.30, 1.00, 1.80, 3.00, 4.50, 5.20, 6.45, 7.00,
7.10, 8.00, 8.03, 8.07, 8.10, 9.00, 9.30, 10.20, 10.50, 10.6])
expected = np.array([0.6, 1.46333333, 2.45, 3.1, 3.81590909,
4.15, 4.35, 4.15483871, 4.11935484, 3.8,
3.79884615, 3.79730769, 3.79615385, 3.76153846, 3.75,
3., 2., 0.9])
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# Reversed case.
spectrum = spectrum[::-1]
bands = np.cumsum(np.concatenate(
(np.array([0.3]), (bands[1:] - bands[:-1])[::-1])))
expected = expected[::-1]
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# A case with two valid local maxima.
spectrum = np.array([0.60, 1.00, 2.45, 3.10, 3.25, 4.15, 4.35, 4.10,
3.10, 3.80, 3.50, 3.75, 2.00, 2.85, 3.70, 3.00, 2.00, 0.90])
bands = np.array([0.30, 1.00, 1.80, 3.00, 4.50, 5.20, 6.45, 7.00,
7.10, 8.00, 8.03, 8.07, 8.10, 9.00, 9.30, 10.20, 10.50, 10.6])
expected = np.array([0.6, 1.46333333, 2.45, 3.1, 3.81590909,
4.15, 4.35, 4.15483871, 4.11935484, 3.8,
3.77857143, 3.75, 3.74878049, 3.71219512, 3.7,
3., 2., 0.9])
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# Reversed case.
spectrum = spectrum[::-1]
bands = np.cumsum(np.concatenate(
(np.array([0.3]), (bands[1:] - bands[:-1])[::-1])))
expected = expected[::-1]
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# A case with two valid local maxima, but one covering eliminating the
# other.
spectrum = np.array([0.60, 1.00, 2.45, 3.10, 3.25, 4.15, 4.35, 4.10,
3.10, 3.80, 3.50, 3.85, 2.00, 2.85, 3.70, 3.00, 2.00, 0.90])
bands = np.array([0.30, 1.00, 1.80, 3.00, 4.50, 5.20, 6.45, 7.00,
7.10, 8.00, 8.03, 8.07, 8.10, 9.00, 9.30, 10.20, 10.50, 10.6])
expected = np.array([0.6, 1.46333333, 2.45, 3.1, 3.81590909,
4.15, 4.35, 4.18024691, 4.14938272, 3.87160494,
3.86234568, 3.85, 3.84634146, 3.73658537, 3.7,
3., 2., 0.9])
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
# Reversed case.
spectrum = spectrum[::-1]
bands = np.cumsum(np.concatenate(
(np.array([0.3]), (bands[1:] - bands[:-1])[::-1])))
expected = expected[::-1]
assert_allclose(expected, spectral_continuum(
spectrum, bands, 'segmented'))
def test_2d_array(self):
part = self.image[20:22, 20:22].reshape(4, 220)
cnt = spectral_continuum(part, self.bands)
# Check some values to make sure results are sane.
assert(cnt[0, 200] == 1422)
assert(cnt[1, 200] == 1421)
assert(cnt[2, 200] == 1469)
assert(cnt[3, 200] == 1491)
def test_3d_array(self):
part = self.image[20:22, 20:22]
cnt = spectral_continuum(part, self.bands)
# Check some values to make sure results are sane.
assert(cnt[0, 0, 200] == 1422)
assert(cnt[0, 1, 200] == 1421)
assert(cnt[1, 0, 200] == 1469)
assert(cnt[1, 1, 200] == 1491)
def test_out_parameter(self):
part = self.image[20:22, 20:22]
out = np.empty_like(part)
cnt = spectral_continuum(part, self.bands, out=out)
assert(cnt is out)
# And just do a quick check if result is sane.
assert(out[1, 1, 200] == 1491)
class FindContinuumPointsTest(ContinuumTest):
'''Tests continuum_points.'''
def test_points_of_real_spectrum(self):
points = continuum_points(self.image[20, 20], self.bands)
assert(np.array_equal(points[0], self.bands[[0, 1, 2, 5, 6, 41, 219]]))
assert(np.array_equal(points[1], np.array(
[3505, 4141, 4516, 4924, 5002, 4712, 1019], dtype=np.int16)))
def test_points_of_real_spectrum_segmented(self):
# This example includes flat local maxima, that span three or more points.
points = continuum_points(self.image[20, 20], self.bands, 'segmented')
expected_result = (
np.array([400.019989, 409.820007, 419.619995, 449.070007, 458.899994,
783.27002, 802.530029, 841.039978, 860.280029, 879.530029,
994.880005, 1014.090027, 1052.48999, 1244.26001, 1273.,
1282.959961, 1541.589966, 1561.439941, 1620.97998, 1630.900024,
2122.780029, 2132.72998, 2172.5, 2212.219971, 2222.149902,
2311.350098, 2331.139893, 2360.810059, 2390.459961, 2400.330078,
2429.949951, 2459.540039, 2498.959961]),
np.array([3505, 4141, 4516, 4924, 5002, 4712, 4578, 4496, 4424, 4423, 3979,
3925, 3801, 3026, 2852, 2817, 2206, 2204, 2136, 2124, 1377, 1376,
1333, 1317, 1312, 1219, 1202, 1162, 1126, 1122, 1100, 1068, 1019],
dtype=np.int16)
)
assert(np.array_equal(points[0], expected_result[0]))
assert(np.array_equal(points[1], expected_result[1]))
class RemoveContinuumTest(ContinuumTest):
'''Tests remove_continuum.'''
def test_simple_case(self):
continuum_removed = np.array([1., 0.6833713, 1., 1., 0.85169744,
1., 1., 0.96830329, 0.73579013, 0.67102681,
0.49967127, 0.74741201, 1., 1., 1., 1.])
bands = np.array([0.30, 1.00, 1.80, 3.00, 4.50, 5.20, 6.45, 7.00, 7.10, 8.00,
8.10, 9.00, 9.30, 10.20, 10.50, 10.6])
spectrum = np.array([0.60, 1.00, 2.45, 3.10, 3.25, 4.15, 4.35, 4.10, 3.10,
2.70, 2.00, 2.85, 3.75, 3.00, 2.00, 0.90])
assert_allclose(continuum_removed, remove_continuum(spectrum, bands))
def test_simple_case_segmented(self):
continuum_removed = np.array([1., 0.6833713, 1., 1., 0.85169744,
1., 1., 0.98680124, 0.75254503, 1.,
0.52684904, 0.75766871, 1., 1., 1.,
1.])
bands = np.array([0.30, 1.00, 1.80, 3.00, 4.50, 5.20, 6.45, 7.00, 7.10, 8.00,
8.10, 9.00, 9.30, 10.20, 10.50, 10.6])
spectrum = np.array([0.60, 1.00, 2.45, 3.10, 3.25, 4.15, 4.35, 4.10, 3.10,
3.80, 2.00, 2.85, 3.75, 3.00, 2.00, 0.90])
assert_allclose(continuum_removed, remove_continuum(
spectrum, bands, mode='segmented'))
def test_in_and_out_same(self):
part = self.image[20:22, 20:22].astype(np.float64)
res = remove_continuum(part, self.bands, out=part)
# Make sure results are sane.
assert(res[1, 1, 200] == 0.8372113957762342)
assert(res is part)
def run():
print('\n' + '-' * 72)
print('Running continuum tests.')
print('-' * 72)
for T in [FindContinuumTest, FindContinuumPointsTest, RemoveContinuumTest]:
T().run()
if __name__ == '__main__':
from spectral.tests.run import parse_args, reset_stats, print_summary
parse_args()
reset_stats()
run()
print_summary()
|
