mirrors/docling
1
0
Fork 0
mirror of https://github.com/DS4SD/docling.git synced 2025-07-26 20:14:47 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix baseocrmodel test issue

Browse Source 
Signed-off-by: William Easton <bill.easton@elastic.co>
This commit is contained in:
William Easton 2025-07-01 10:43:30 -05:00
parent a2eafdbac3
commit bab9c25c0c
No known key found for this signature in database
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
docling/models/base_ocr_model.py
View File

@ -20,9 +20,6 @@ from docling.models.base_model import BaseModelWithOptions, BasePageModel
_log = logging.getLogger(__name__) _log = logging.getLogger(__name__)
if TYPE_CHECKING:
from scipy.ndimage import binary_dilation, find_objects, label
class BaseOcrModel(BasePageModel, BaseModelWithOptions): class BaseOcrModel(BasePageModel, BaseModelWithOptions):
def __init__( def __init__(
@ -33,6 +30,7 @@ class BaseOcrModel(BasePageModel, BaseModelWithOptions):
options: OcrOptions, options: OcrOptions,
accelerator_options: AcceleratorOptions, accelerator_options: AcceleratorOptions,
): ):
# Make sure any delay/error from import occurs on ocr model init and not first use
from scipy.ndimage import binary_dilation, find_objects, label from scipy.ndimage import binary_dilation, find_objects, label
self.enabled = enabled self.enabled = enabled
@ -40,6 +38,8 @@ class BaseOcrModel(BasePageModel, BaseModelWithOptions):
# Computes the optimum amount and coordinates of rectangles to OCR on a given page # Computes the optimum amount and coordinates of rectangles to OCR on a given page
def get_ocr_rects(self, page: Page) -> List[BoundingBox]: def get_ocr_rects(self, page: Page) -> List[BoundingBox]:
from scipy.ndimage import binary_dilation, find_objects, label
BITMAP_COVERAGE_TRESHOLD = 0.75 BITMAP_COVERAGE_TRESHOLD = 0.75
assert page.size is not None assert page.size is not None