def test_wrong_input_type():
with pytest.raises(AttributeError):
gaussian_blur(12345,
"InstaF_Python/test_py/test_image/test.png",
sigma=1,
filter_shape=(3, 3))
def test_input_path_not_exist():
with pytest.raises(FileNotFoundError):
gaussian_blur("./1234/123.png",
"InstaF_Python/test_py/test_image/test.png",
sigma=1,
filter_shape=(3, 3))
import numpy as np
import pandas as pd
import os
import pytest
import skimage
from skimage.transform import resize
from InstaF_Python.gaussian_blur import gaussian_blur
# input for test one, see if the function is able to correctly blur an ordinary image
# We use Milad's photo as an test example
# to test the output is now defined as a matrix, just to make the testing easier, in the future, the output will be converted to a image
# first we define a function to process the image
output = gaussian_blur("InstaF_Python/test/test_image/carnovsky_RGB_1.png",
"InstaF_Python/test/test_image/output.png",
sigma=1,
filter_shape=(3, 3))
exp_output = skimage.io.imread("InstaF_Python/test/test_image/test.png")
# the first part we want to test if our function is able to convert regular RBG channel image
# we can then calculate the expected output image by matrix multiplication in python by hand for sigma = 1, filter_shape=(3,3)
# test normal picture with RBG channel
def test_normal_pic():
assert np.array_equal(output,
exp_output), "GaussianBlur not working on RBG image"
# test non-image input