-
Notifications
You must be signed in to change notification settings - Fork 0
/
test_pyautocv.py
150 lines (105 loc) · 5.85 KB
/
test_pyautocv.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
import pytest
from pyautocv.segmentation import *
import os
from unittest import mock
dir_path = os.path.dirname(os.path.abspath(__file__))
def make_path(use_path):
return os.path.join(dir_path, use_path)
@pytest.fixture
def test_object():
def _test_object(dir_path, image_suffix = "png", **kwargs):
return Segmentation(directory = make_path(dir_path), image_suffix= image_suffix, **kwargs)
return _test_object
@pytest.fixture
def use_object(test_object):
def callable_obj():
return test_object(dir_path="images/cats")
return callable_obj
def assert_instance(use_object):
# Very basic assertion. For whatever reason we fail to find tests when using parametrize
assert isinstance(use_object(), Segmentation)
def test_not_object(test_object):
with pytest.raises(NotADirectoryError, match = "not a directory is not a valid directory in the current path.") as err:
test_object(dir_path = "not a directory", image_suffix = "png")
def test_error_throws(test_object):
with pytest.raises(ValueError, match = "Only png, jpg, and tif are supported" ) as err:
test_object(dir_path = "images/cars", image_suffix = "nope")
# TODO: Use a recursive fixture for this
def test_graying(use_object):
grayed = gray_images(use_object().read_images())
assert len(use_object().read_images()[0].shape) == 3
# drop extra color index after graying for the first image
assert len(grayed[0].shape) == 2
# No fixture needed here
def test_gray_fails():
with pytest.raises(TypeError, match = "Expected a list of images not str") as err:
gray_images("not a list")
def test_smoothing(test_object, use_object):
with pytest.raises(ValueError, match = "mask should be one of mean, median, box, and gaussian") as err:
use_object().smooth(mask="nope")
with pytest.raises(TypeError, match = "Expected a tuple not int") as err:
use_object().smooth(mask = "mean", kernel_shape = 5)
# Expect a list if all works as expected
smoothed = test_object(dir_path = "images/dic", image_suffix = "tif",
color_mode="gray").smooth(mask="median")
# TODO: use more meaningful tests not just a list instance
assert isinstance(smoothed, list)
def test_thresholding(use_object):
with pytest.raises(ValueError, match = "Thresholding with nope is not supported") as err:
use_object().threshold_images(threshold_method="nope")
assert use_object().read_images()[1].flat[40] != use_object().threshold_images()[1].flat[40]
def test_edge_detection(test_object, use_object):
with pytest.raises(ValueError, match = "Edge detection with gibberish not supported." ) as err:
use_object().detect_edges(operator="gibberish")
# If edge detecting, then we expect non equality with read images
# This is not very pretty and there might be a better way, use this for now
assert use_object().read_images()[1].flat[40] != use_object().detect_edges()[1].flat[40]
# Test gray color mode
use_gray = Segmentation(directory = make_path("images/dic"), image_suffix="tif", color_mode="gray")
assert use_gray.read_images()[1].flat[40] != use_gray.detect_edges()[1].flat[40]
# Test that if we provide sub-folders we get the expected image length
assert len(use_object().read_images(other_directory = make_path("images/cats"))) == 6
def test_image_reading(test_object, use_object):
# Expect length 3 since we have two jpg and one png in images/car
assert len(use_object().read_images()) == 3
# Should use pil, expect 15
tif_only = test_object(dir_path = "images/dic", image_suffix="tif")
assert len(tif_only.read_images()) == 15
def test_resizing(use_object):
resized = resize_images(use_object().read_images(), (125, 125))
assert resized[0].shape[0] == 125
with pytest.raises(TypeError, match ="Expected a tuple in target_size not int") as err:
resize_images(use_object().read_images(), 4)
with pytest.raises(ValueError, match = "Please provide both an image list and a target size") as err:
resize_images()
def test_reshaping(use_object):
with pytest.raises(ValueError, match ="Please provide a list of images to reshape.") as err:
reshape_images()
# Check that reshaped images are not the same shape as original images
reshaped = reshape_images(use_object().read_images())
assert reshaped[0].shape == use_object().read_images()[0].shape
def test_stacking(use_object):
with pytest.raises(ValueError, match = "Please provide two lists to stack") as err:
stack_images()
with pytest.raises(TypeError, match ="Both list_one and list_two should be lists" ) as err:
stack_images(use_object().read_images(), "gibberish")
with pytest.raises(ValueError, match = "direction should be one of horizontal, vertical, h, v not gibberish") as err:
stack_images(use_object().read_images(), use_object().threshold_images(),
direction="gibberish")
assert len(stack_images(use_object().read_images(), use_object().read_images())) == 3
def test_hist_plots(use_object):
# use_object = test_object(dir_path="images/cats")
# This is just a simple workaround as it seems we cannot actually
# use the decorator @mock.patch within the pytest framework.
with mock.patch("pyautocv.segmentation.plt") as mock_plt:
image_test = use_object().read_images()[1]
plot_hist(image_test)
mock_plt.plot.assert_called_once()
# TODO assert that color_mode works as expected
mock_plt.xlim.assert_called_once_with([0, 256])
# TODO: Figure out why this only fails as a test.
@pytest.mark.xfail
def test_show_images(use_object, mock_plt):
with mock.patch("pyautocv.segmentation.plt") as mock_plt:
show_images(use_object.read_images(), use_object.read_images())
mock_plt.subplots.assert_called_once()