image-statistics-matching/tests/matching/operations/test_histogram_matching.py at master · continental/image-statistics-matching · GitHub

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
# pylint: disable=missing-module-docstring (C0114)
# pylint: disable=missing-function-docstring (C0116)
import inspect

import cv2
import numpy as np
import pytest
from skimage.exposure import match_histograms

from core import DIM_3, MATCH_FULL, MATCH_ZERO
from matching import ChannelsType
from matching.operations import HistogramMatching
from tests import CHANNELS_DEFAULT, MUNICH_1_PATH, MUNICH_2_PATH

from . import (TEST_REF_IMAGE, TEST_RES_IMAGE_02, TEST_RES_IMAGE_05,
               TEST_RES_IMAGE_08, TEST_SRC_IMAGE)

# ground truth result image with matching proportion 1.0
TEST_RES_IMAGE = match_histograms(TEST_SRC_IMAGE, TEST_REF_IMAGE)


@pytest.fixture(name='hist_match')
def fixture_histogram_matching() -> HistogramMatching:
    return HistogramMatching(CHANNELS_DEFAULT, check_input=True)


def test_design() -> None:
    assert inspect.isabstract(HistogramMatching) is False
    assert len(HistogramMatching.__mro__) == 4


@pytest.mark.parametrize('match_prop', [list(), (1, 2, 3)])
def test_hm_match_prop_type_error(match_prop: float) -> None:
    with pytest.raises(TypeError):
        HistogramMatching(CHANNELS_DEFAULT, check_input=True,
                          match_prop=match_prop)


@pytest.mark.parametrize('match_prop', ['abc'])
def test_hm_match_prop_value_error(match_prop: float) -> None:
    with pytest.raises(ValueError):
        HistogramMatching(CHANNELS_DEFAULT, check_input=True,
                          match_prop=match_prop)


@pytest.mark.parametrize('match_prop', [0, False, True, 1, '0.5', '1.0'])
def test_hm_match_prop(match_prop: float) -> None:
    hist_match = HistogramMatching(CHANNELS_DEFAULT, check_input=True,
                                   match_prop=match_prop)
    assert hist_match.match_prop == float(match_prop)


@pytest.mark.parametrize('match_prop',
                         [MATCH_ZERO, 0.2, 0.6, MATCH_FULL, '0.3', 1])
def test_hm_match_prop_valid_value(match_prop: float) -> None:
    hist_match = HistogramMatching(CHANNELS_DEFAULT, check_input=True,
                                   match_prop=match_prop)
    assert hist_match.match_prop == float(match_prop)


@pytest.mark.parametrize('match_prop', [-5., -0.2, 1.1, 8.0, '-10', '4.5', 14])
def test_hm_match_prop_invalid_value(match_prop: float) -> None:
    with pytest.raises(ValueError):
        HistogramMatching(CHANNELS_DEFAULT, check_input=True,
                          match_prop=match_prop)


def test_apply(hist_match: HistogramMatching) -> None:
    # pylint: disable=protected-access (W0212)
    hist_match.channels = (0,)
    source = TEST_SRC_IMAGE[:, :, np.newaxis].astype(float)
    reference = TEST_REF_IMAGE[:, :, np.newaxis].astype(float)
    result = hist_match(source, reference)
    expected_result = hist_match._apply(source, reference)
    np.testing.assert_array_equal(result, expected_result)


def test_match_channel(hist_match: HistogramMatching) -> None:
    # pylint: disable=protected-access (W0212)
    result = hist_match._match_channel(TEST_SRC_IMAGE, TEST_REF_IMAGE)

    # we test against scikit image histogram matching
    assert result.shape == TEST_SRC_IMAGE.shape
    np.testing.assert_array_equal(result, TEST_RES_IMAGE)


@pytest.mark.parametrize('match_prop, expected_result',
                         [(MATCH_ZERO, TEST_SRC_IMAGE),
                          (0.2, TEST_RES_IMAGE_02),
                          (0.5, TEST_RES_IMAGE_05),
                          (0.8, TEST_RES_IMAGE_08),
                          (MATCH_FULL, TEST_RES_IMAGE)])
def test_match_channel_prop(match_prop: float,
                            expected_result: np.ndarray) -> None:
    # pylint: disable=protected-access (W0212)
    hist_match = HistogramMatching(CHANNELS_DEFAULT, check_input=True,
                                   match_prop=match_prop)
    result = hist_match._match_channel(TEST_SRC_IMAGE, TEST_REF_IMAGE)

    assert result.shape == TEST_SRC_IMAGE.shape
    np.testing.assert_array_almost_equal(result, expected_result)


@pytest.mark.parametrize('source_path, reference_path',
                         [(MUNICH_1_PATH, MUNICH_2_PATH),
                          (MUNICH_2_PATH, MUNICH_1_PATH),
                          (MUNICH_1_PATH, MUNICH_1_PATH)])
def test_match_channel_images(source_path: str, reference_path: str,
                              hist_match: HistogramMatching) -> None:
    # pylint: disable=protected-access (W0212)
    source = cv2.imread(source_path)
    reference = cv2.imread(reference_path)

    for channel in range(source.shape[-1]):
        source_c = source[:, :, channel]
        reference_c = reference[:, :, channel]
        result = hist_match._match_channel(source_c, reference_c)

        # we test against scikit image histogram matching
        expected_result = match_histograms(source_c, reference_c)

        assert result.shape == source_c.shape
        np.testing.assert_array_equal(result, expected_result)


@pytest.mark.parametrize('channels',
                         [CHANNELS_DEFAULT, (0, 1), (1,), (0, 2), (1, 2)])
def test_apply_channels(channels: ChannelsType) -> None:
    source = cv2.imread(MUNICH_1_PATH)
    reference = cv2.imread(MUNICH_2_PATH)

    original_source = np.copy(source)
    original_reference = np.copy(reference)

    hist_match = HistogramMatching(channels, check_input=True)
    result = hist_match(source.astype(float), reference.astype(float))

    # check channels to be matched
    for channel in channels:
        with np.testing.assert_raises(AssertionError):
            np.testing.assert_array_equal(source[:, :, channel],
                                          result[:, :, channel])

    # check skipped channels
    skipped_channels = tuple(set(CHANNELS_DEFAULT) - set(channels))
    for channel in skipped_channels:
        np.testing.assert_array_equal(source[:, :, channel],
                                      result[:, :, channel])

    assert result.shape == source.shape
    assert result.dtype == np.float32

    np.testing.assert_array_equal(source, original_source)
    np.testing.assert_array_equal(reference, original_reference)


def test_apply_2d_image(hist_match: HistogramMatching) -> None:
    hist_match.channels = (0,)
    source = TEST_SRC_IMAGE[:, :, np.newaxis]
    reference = TEST_REF_IMAGE[:, :, np.newaxis]
    original_source = np.copy(source)
    original_reference = np.copy(reference)

    result = hist_match(source.astype(float), reference.astype(float))

    assert result.shape == original_source.shape
    assert result.dtype == np.float32
    assert result.ndim == DIM_3

    np.testing.assert_array_equal(source, original_source)
    np.testing.assert_array_equal(reference, original_reference)

    # we test against scikit image histogram matching
    np.testing.assert_array_equal(result, TEST_RES_IMAGE[:, :, np.newaxis])