Inspect picture content

In [ ]:

from docling_core.types.doc import TextItem

from docling_core.types.doc import TextItem

In [ ]:

from docling.datamodel.base_models import InputFormat
from docling.datamodel.pipeline_options import PdfPipelineOptions
from docling.document_converter import DocumentConverter, PdfFormatOption

from docling.datamodel.base_models import InputFormat from docling.datamodel.pipeline_options import PdfPipelineOptions from docling.document_converter import DocumentConverter, PdfFormatOption

In [ ]:

source = "tests/data/pdf/amt_handbook_sample.pdf"

source = "tests/data/pdf/amt_handbook_sample.pdf"

In [ ]:

pipeline_options = PdfPipelineOptions()
pipeline_options.images_scale = 2
pipeline_options.generate_page_images = True

pipeline_options = PdfPipelineOptions() pipeline_options.images_scale = 2 pipeline_options.generate_page_images = True

In [ ]:

doc_converter = DocumentConverter(
    format_options={InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)}
)

doc_converter = DocumentConverter( format_options={InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)} )

In [ ]:

result = doc_converter.convert(source)

result = doc_converter.convert(source)

In [ ]:

doc = result.document

doc = result.document

In [ ]:

for picture in doc.pictures:
    # picture.get_image(doc).show() # display the picture
    print(picture.caption_text(doc), " contains these elements:")

    for item, level in doc.iterate_items(root=picture, traverse_pictures=True):
        if isinstance(item, TextItem):
            print(item.text)

    print("\n")

for picture in doc.pictures: # picture.get_image(doc).show() # display the picture print(picture.caption_text(doc), " contains these elements:") for item, level in doc.iterate_items(root=picture, traverse_pictures=True): if isinstance(item, TextItem): print(item.text) print("\n")