import cv2 import numpy as np import pytest from cv_analysis.figure_detection.text import ( remove_primary_text_regions, apply_threshold_to_image, ) from test.utils.utils import powerset @pytest.mark.parametrize("error_tolerance", [0.07]) @pytest.mark.parametrize("background_color", [255, 220]) class TestFindPrimaryTextRegions: def test_blank_page_stays_blank(self, background, error_tolerance): result_page = remove_primary_text_regions(background) np.testing.assert_equal(result_page, apply_threshold_to_image(background)) @pytest.mark.parametrize("image_size", [(200, 200), (500, 500), (800, 800)]) def test_page_without_text_keeps_images(self, page_with_images, error_tolerance): result_page = remove_primary_text_regions(page_with_images) np.testing.assert_equal(result_page, apply_threshold_to_image(page_with_images)) @pytest.mark.parametrize("font_scale", [1, 1.5, 2]) @pytest.mark.parametrize("font_style", [cv2.FONT_HERSHEY_SIMPLEX, cv2.FONT_HERSHEY_COMPLEX]) @pytest.mark.parametrize("text_types", powerset(["body", "header", "caption"])) def test_page_with_only_text_gets_text_removed(self, page_with_text, error_tolerance): result_page = remove_primary_text_regions(page_with_text) relative_error = np.sum(result_page != apply_threshold_to_image(page_with_text)) / result_page.size assert relative_error <= error_tolerance @pytest.mark.parametrize("image_size", [(200, 200), (500, 500), (800, 800)]) @pytest.mark.parametrize("font_scale", [1, 1.5, 2]) @pytest.mark.parametrize("font_style", [cv2.FONT_HERSHEY_SIMPLEX, cv2.FONT_HERSHEY_COMPLEX]) @pytest.mark.parametrize("text_types", powerset(["body", "header", "caption"])) def test_page_with_images_and_text_keeps_images(self, page_with_images_and_text, error_tolerance): result_page = remove_primary_text_regions(page_with_images_and_text) relative_error = np.sum(result_page != apply_threshold_to_image(page_with_images_and_text)) / result_page.size assert relative_error <= error_tolerance