a871fa3bd3
Merge in RR/cv-analysis from refactor-evaluate to master
Squashed commit of the following:
commit cde03a492452610322f8b7d3eb804a51afb76d81
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Fri Jul 22 12:37:36 2022 +0200
add optional show analysis metadata dict
commit fb8bb9e2afa7767f2560f865516295be65f97f20
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Fri Jul 22 12:13:18 2022 +0200
add script to evaluate runtime per page for all cv-analysis operations for multiple PDFs
commit 721e823e2ec38aae3fea51d01e2135fc8f228d94
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Fri Jul 22 10:30:31 2022 +0200
refactor
commit a453753cfa477e162e5902ce191ded61cb678337
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Fri Jul 22 10:19:24 2022 +0200
add logic to transform result coordinates accordingly to page rotation, update annotation script to use this logic
commit 71c09758d0fb763a2c38c6871e1d9bf51f2e7c41
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Thu Jul 21 15:57:49 2022 +0200
introduce pipeline for image conversion, analysis and result formatting
commit aef252a41b9658dd0c4f55aa2d9f84de933586e0
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Thu Jul 21 15:57:38 2022 +0200
introduce pipeline for image conversion, analysis and result formatting
25 lines
747 B
Python
25 lines
747 B
Python
import fitz
|
|
import numpy as np
|
|
import pytest
|
|
|
|
from cv_analysis.utils.pdf2image import pdf_to_image_metadata_pairs
|
|
|
|
|
|
@pytest.fixture
|
|
def pdf(n_pages):
|
|
doc = fitz.open()
|
|
for n in range(n_pages):
|
|
page = doc.new_page()
|
|
where = fitz.Point(50, 100)
|
|
page.insert_text(where, "De gustibus non est disputandum.", fontsize=30)
|
|
return doc.write()
|
|
|
|
|
|
@pytest.mark.parametrize("n_pages", [1])
|
|
def test_pdf_to_array_and_metadata(pdf):
|
|
for image_metadata_pair in pdf_to_image_metadata_pairs(pdf):
|
|
assert isinstance(image_metadata_pair.image, np.ndarray)
|
|
assert image_metadata_pair.image.shape == (2339, 1653, 3) # Height, Width, Color channels
|
|
|
|
assert isinstance(image_metadata_pair.metadata, dict)
|