mirror of
https://github.com/DS4SD/docling.git
synced 2025-07-23 18:45:00 +00:00
fix(ocr-utils): unit test and fix the rotate_bounding_box
function (#1897)
Signed-off-by: Clément Doumouro <clement.doumouro@gmail.com>
This commit is contained in:
parent
a07ba863c4
commit
931eb55b88
@ -41,7 +41,7 @@ def tesseract_box_to_bounding_rectangle(
|
||||
im_size: Tuple[int, int],
|
||||
) -> BoundingRectangle:
|
||||
# box is in the top, left, height, width format, top left coordinates
|
||||
rect = rotate_bounding_box(bbox, angle=-orientation, im_size=im_size)
|
||||
rect = rotate_bounding_box(bbox, angle=orientation, im_size=im_size)
|
||||
rect = BoundingRectangle(
|
||||
r_x0=rect.r_x0 / scale,
|
||||
r_y0=rect.r_y0 / scale,
|
||||
|
@ -14,43 +14,36 @@ def rotate_bounding_box(
|
||||
# coordinate system. Then other corners are found rotating counterclockwise
|
||||
bbox = bbox.to_top_left_origin(im_size[1])
|
||||
left, top, width, height = bbox.l, bbox.t, bbox.width, bbox.height
|
||||
im_h, im_w = im_size
|
||||
im_w, im_h = im_size
|
||||
angle = angle % 360
|
||||
if angle == 0:
|
||||
r_x0 = left
|
||||
r_y0 = top + height
|
||||
r_x1 = r_x0 + width
|
||||
r_y1 = r_y0
|
||||
r_x2 = r_x0 + width
|
||||
r_y2 = r_y0 - height
|
||||
r_x3 = r_x0
|
||||
r_y3 = r_y0 - height
|
||||
return BoundingRectangle.from_bounding_box(bbox)
|
||||
elif angle == 90:
|
||||
r_x0 = im_w - (top + height)
|
||||
r_x0 = top + height
|
||||
r_y0 = im_w - left
|
||||
r_x1 = r_x0
|
||||
r_y1 = r_y0 - width
|
||||
r_x2 = r_x1 - height
|
||||
r_y2 = r_y1
|
||||
r_x3 = r_x2
|
||||
r_y3 = r_y0
|
||||
elif angle == 180:
|
||||
r_x0 = im_w - left
|
||||
r_y0 = im_h - (top + height)
|
||||
r_x1 = r_x0 - width
|
||||
r_y1 = r_y0
|
||||
r_x2 = r_x1
|
||||
r_y2 = r_y1 + height
|
||||
r_x3 = r_x0
|
||||
r_y3 = r_y2
|
||||
elif angle == 270:
|
||||
r_x0 = im_h - (top + height)
|
||||
r_y0 = left
|
||||
r_x1 = r_x0
|
||||
r_y1 = r_y0 + width
|
||||
r_x2 = r_x0 + height
|
||||
r_y2 = r_y0 + width
|
||||
r_x3 = r_x0
|
||||
r_y3 = r_y0 + width
|
||||
elif angle == 180:
|
||||
r_x0 = im_h - left
|
||||
r_y0 = im_w - (top + height)
|
||||
r_x1 = r_x0 - width
|
||||
r_y1 = r_y0
|
||||
r_x2 = r_x0 - width
|
||||
r_y2 = r_y0 + height
|
||||
r_x3 = r_x0
|
||||
r_y3 = r_y0 + height
|
||||
elif angle == 270:
|
||||
r_x0 = top + height
|
||||
r_y0 = im_h - left
|
||||
r_x1 = r_x0
|
||||
r_y1 = r_y0 - width
|
||||
r_x2 = r_x0 - height
|
||||
r_y2 = r_y0 - width
|
||||
r_x3 = r_x0 - height
|
||||
r_x2 = r_x1 + height
|
||||
r_y2 = r_y1
|
||||
r_x3 = r_x2
|
||||
r_y3 = r_y0
|
||||
else:
|
||||
msg = (
|
||||
@ -58,7 +51,7 @@ def rotate_bounding_box(
|
||||
f" {sorted(CLIPPED_ORIENTATIONS)}"
|
||||
)
|
||||
raise ValueError(msg)
|
||||
return BoundingRectangle(
|
||||
rectangle = BoundingRectangle(
|
||||
r_x0=r_x0,
|
||||
r_y0=r_y0,
|
||||
r_x1=r_x1,
|
||||
@ -69,3 +62,4 @@ def rotate_bounding_box(
|
||||
r_y3=r_y3,
|
||||
coord_origin=CoordOrigin.TOPLEFT,
|
||||
)
|
||||
return rectangle
|
||||
|
@ -44,7 +44,7 @@ authors = [
|
||||
requires-python = '>=3.9,<4.0'
|
||||
dependencies = [
|
||||
'pydantic (>=2.0.0,<3.0.0)',
|
||||
'docling-core[chunking] (>=2.39.0,<3.0.0)',
|
||||
'docling-core[chunking] (>=2.40.0,<3.0.0)',
|
||||
'docling-parse (>=4.0.0,<5.0.0)',
|
||||
'docling-ibm-models (>=3.6.0,<4)',
|
||||
'filetype (>=1.2.0,<2.0.0)',
|
||||
|
80
tests/test_ocr_utils.py
Normal file
80
tests/test_ocr_utils.py
Normal file
@ -0,0 +1,80 @@
|
||||
from typing import Tuple
|
||||
|
||||
import pytest
|
||||
from docling_core.types.doc import BoundingBox, CoordOrigin
|
||||
from docling_core.types.doc.page import BoundingRectangle
|
||||
|
||||
from docling.utils.orientation import rotate_bounding_box
|
||||
|
||||
IM_SIZE = (4, 5)
|
||||
BBOX = BoundingBox(l=1, t=3, r=3, b=4, coord_origin=CoordOrigin.TOPLEFT)
|
||||
RECT = BoundingRectangle(
|
||||
r_x0=1,
|
||||
r_y0=4,
|
||||
r_x1=3,
|
||||
r_y1=4,
|
||||
r_x2=3,
|
||||
r_y2=3,
|
||||
r_x3=1,
|
||||
r_y3=3,
|
||||
coord_origin=CoordOrigin.TOPLEFT,
|
||||
)
|
||||
RECT_90 = BoundingRectangle(
|
||||
r_x0=4,
|
||||
r_y0=3,
|
||||
r_x1=4,
|
||||
r_y1=1,
|
||||
r_x2=3,
|
||||
r_y2=1,
|
||||
r_x3=3,
|
||||
r_y3=3,
|
||||
coord_origin=CoordOrigin.TOPLEFT,
|
||||
)
|
||||
RECT_180 = BoundingRectangle(
|
||||
r_x0=3,
|
||||
r_y0=1,
|
||||
r_x1=1,
|
||||
r_y1=1,
|
||||
r_x2=1,
|
||||
r_y2=2,
|
||||
r_x3=3,
|
||||
r_y3=2,
|
||||
coord_origin=CoordOrigin.TOPLEFT,
|
||||
)
|
||||
RECT_270 = BoundingRectangle(
|
||||
r_x0=1,
|
||||
r_y0=1,
|
||||
r_x1=1,
|
||||
r_y1=3,
|
||||
r_x2=2,
|
||||
r_y2=3,
|
||||
r_x3=2,
|
||||
r_y3=1,
|
||||
coord_origin=CoordOrigin.TOPLEFT,
|
||||
)
|
||||
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
["bbox", "im_size", "angle", "expected_rectangle"],
|
||||
[
|
||||
# (BBOX, IM_SIZE, 0, RECT),
|
||||
# (BBOX, IM_SIZE, 90, RECT_90),
|
||||
(BBOX, IM_SIZE, 180, RECT_180),
|
||||
# (BBOX, IM_SIZE, 270, RECT_270),
|
||||
# (BBOX, IM_SIZE, 360, RECT),
|
||||
# (BBOX, IM_SIZE, -90, RECT_270),
|
||||
(BBOX, IM_SIZE, -180, RECT_180),
|
||||
# (BBOX, IM_SIZE, -270, RECT_90),
|
||||
],
|
||||
)
|
||||
def test_rotate_bounding_box(
|
||||
bbox: BoundingBox,
|
||||
im_size: Tuple[int, int],
|
||||
angle: int,
|
||||
expected_rectangle: BoundingRectangle,
|
||||
):
|
||||
rotated = rotate_bounding_box(bbox, angle, im_size)
|
||||
|
||||
assert rotated == expected_rectangle
|
||||
expected_angle_360 = angle % 360
|
||||
assert rotated.angle_360 == expected_angle_360
|
8
uv.lock
generated
8
uv.lock
generated
@ -904,7 +904,7 @@ requires-dist = [
|
||||
{ name = "accelerate", marker = "extra == 'vlm'", specifier = ">=1.2.1,<2.0.0" },
|
||||
{ name = "beautifulsoup4", specifier = ">=4.12.3,<5.0.0" },
|
||||
{ name = "certifi", specifier = ">=2024.7.4" },
|
||||
{ name = "docling-core", extras = ["chunking"], specifier = ">=2.39.0,<3.0.0" },
|
||||
{ name = "docling-core", extras = ["chunking"], specifier = ">=2.40.0,<3.0.0" },
|
||||
{ name = "docling-ibm-models", specifier = ">=3.6.0,<4" },
|
||||
{ name = "docling-parse", specifier = ">=4.0.0,<5.0.0" },
|
||||
{ name = "easyocr", specifier = ">=1.7,<2.0" },
|
||||
@ -979,7 +979,7 @@ examples = [
|
||||
|
||||
[[package]]
|
||||
name = "docling-core"
|
||||
version = "2.39.0"
|
||||
version = "2.40.0"
|
||||
source = { registry = "https://pypi.org/simple" }
|
||||
dependencies = [
|
||||
{ name = "jsonref" },
|
||||
@ -993,9 +993,9 @@ dependencies = [
|
||||
{ name = "typer" },
|
||||
{ name = "typing-extensions" },
|
||||
]
|
||||
sdist = { url = "https://files.pythonhosted.org/packages/6a/8b/5613467523bed58d9f2b94220947783914b6d9910a8d20908cf148805427/docling_core-2.39.0.tar.gz", hash = "sha256:77530156c79c9000fe3104894935437d3e2d46dc0f567b5a500974d7c1a8b38b", size = 148005, upload-time = "2025-06-27T12:59:56.694Z" }
|
||||
sdist = { url = "https://files.pythonhosted.org/packages/02/5d/fb9fc563d694259877a94f9ae7cf77eba4e1143e539a9dda9fc738db1548/docling_core-2.40.0.tar.gz", hash = "sha256:80a03ac0869d45e1b15ac122ed9da1951cb8d209f596269042601d42e4e1f47f", size = 148373, upload-time = "2025-07-02T16:23:23.754Z" }
|
||||
wheels = [
|
||||
{ url = "https://files.pythonhosted.org/packages/70/85/3d59ac46a47f62a0ed79e187c4163cecd2693d05006f771038db4781f9ff/docling_core-2.39.0-py3-none-any.whl", hash = "sha256:b7ce5142ab95bd8d5cfe5d7df167a96a6eb41d884f00ea42bb3dd8f40ade92ea", size = 152890, upload-time = "2025-06-27T12:59:55.327Z" },
|
||||
{ url = "https://files.pythonhosted.org/packages/c9/5c/a66db0c0724a0dc7d683e0cd45c90a3f46273f5379f63bdd29152a724061/docling_core-2.40.0-py3-none-any.whl", hash = "sha256:439ae2aab3a2e4044df9ae76926325f8ae65dac6b3e6fd1911168cbdf6df27df", size = 153028, upload-time = "2025-07-02T16:23:22.064Z" },
|
||||
]
|
||||
|
||||
[package.optional-dependencies]
|
||||
|
Loading…
Reference in New Issue
Block a user