fix generation of images and adapt examples

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

docling/datamodel/document.py

@@ -437,25 +437,6 @@ class ConversionResult(BaseModel):
 
         return ds_doc
 
-    def render_element_images(
-        self, element_types: Tuple[Type[PageElement]] = (FigureElement,)
-    ):
-        for element in self.assembled.elements:
-            if isinstance(element, element_types):
-                page_ix = element.page_no
-                page = self.pages[page_ix]
-
-                assert page.size is not None
-
-                scale = page._default_image_scale
-                crop_bbox = element.cluster.bbox.scaled(scale=scale).to_top_left_origin(
-                    page_height=page.size.height * scale
-                )
-                page_img = page.image
-                if page_img is not None:
-                    cropped_im = page_img.crop(crop_bbox.as_tuple())
-                    yield element, cropped_im
-
 
 class _DocumentConversionInput(BaseModel):
 

docling/pipeline/standard_pdf_pipeline.py

@@ -152,8 +152,8 @@ class StandardPdfPipeline(PaginatedPipeline):
         if self.pipeline_options.generate_page_images:
             for page in conv_res.pages:
                 assert page.image is not None
-                page_ix = page.page_no - 1
+                page_no = page.page_no + 1
-                conv_res.document.pages[page_ix].image = ImageRef.from_pil(
+                conv_res.document.pages[page_no].image = ImageRef.from_pil(
                     page.image, dpi=int(72 * self.pipeline_options.images_scale)
                 )
 
@@ -174,17 +174,17 @@ class StandardPdfPipeline(PaginatedPipeline):
                     and self.pipeline_options.generate_table_images
                 ):
                     page_ix = element.prov[0].page_no - 1
+                    page = conv_res.pages[page_ix]
+                    assert page.size is not None
+                    assert page.image is not None
+
                     crop_bbox = (
                         element.prov[0]
                         .bbox.scaled(scale=scale)
-                        .to_top_left_origin(
+                        .to_top_left_origin(page_height=page.size.height * scale)
-                            page_height=conv_res.pages[page_ix].size.height * scale
-                        )
                     )
 
-                    cropped_im = conv_res.pages[page_ix].image.crop(
+                    cropped_im = page.image.crop(crop_bbox.as_tuple())
-                        crop_bbox.as_tuple()
-                    )
                     element.image = ImageRef.from_pil(cropped_im, dpi=int(72 * scale))
 
         return conv_res

docs/examples/develop_picture_enrichment.py

@@ -34,7 +34,7 @@ class ExamplePictureClassifierEnrichmentModel(BaseEnrichmentModel):
             assert isinstance(element, PictureItem)
 
             # uncomment this to interactively visualize the image
-            element.image.pil_image.show()
+            # element.image.pil_image.show()
 
             element.data.classification = PictureClassificationData(
                 provenance="example_classifier-0.0.1",

docs/examples/export_figures.py

@@ -1,7 +1,8 @@
 import logging
+import time
 from pathlib import Path
 
-import time
+from docling_core.types.doc.document import PictureItem, TableItem
 
 from docling.datamodel.base_models import FigureElement, InputFormat, Table
 from docling.datamodel.pipeline_options import PdfPipelineOptions
@@ -20,10 +21,15 @@ def main():
 
     # Important: For operating with page images, we must keep them, otherwise the DocumentConverter
     # will destroy them for cleaning up memory.
-    # This is done by setting AssembleOptions.images_scale, which also defines the scale of images.
+    # This is done by setting PdfPipelineOptions.images_scale, which also defines the scale of images.
     # scale=1 correspond of a standard 72 DPI image
+    # The PdfPipelineOptions.generate_* are the selectors for the document elements which will be enriched
+    # with the image field
     pipeline_options = PdfPipelineOptions()
     pipeline_options.images_scale = IMAGE_RESOLUTION_SCALE
+    pipeline_options.generate_page_images = True
+    pipeline_options.generate_table_images = True
+    pipeline_options.generate_picture_images = True
 
     doc_converter = DocumentConverter(
         format_options={
@@ -38,20 +44,32 @@ def main():
     output_dir.mkdir(parents=True, exist_ok=True)
     doc_filename = conv_res.input.file.stem
 
-    # Export page images
+    # Save page images
-    for page in conv_res.pages:
+    for page_no, page in conv_res.document.pages.items():
-        page_no = page.page_no + 1
+        page_no = page.page_no
         page_image_filename = output_dir / f"{doc_filename}-{page_no}.png"
         with page_image_filename.open("wb") as fp:
-            page.image.save(fp, format="PNG")
+            page.image.pil_image.save(fp, format="PNG")
 
-    # Export figures and tables
+    # Save images of figures and tables
-    for element, image in conv_res.render_element_images(
+    table_counter = 0
-        element_types=(FigureElement, Table)
+    picture_counter = 0
-    ):
+    for element, _level in conv_res.document.iterate_items():
-        element_image_filename = output_dir / f"{doc_filename}-element-{element.id}.png"
+        if isinstance(element, TableItem):
-        with element_image_filename.open("wb") as fp:
+            table_counter += 1
-            image.save(fp, "PNG")
+            element_image_filename = (
+                output_dir / f"{doc_filename}-table-{table_counter}.png"
+            )
+            with element_image_filename.open("wb") as fp:
+                element.image.pil_image.save(fp, "PNG")
+
+        if isinstance(element, PictureItem):
+            picture_counter += 1
+            element_image_filename = (
+                output_dir / f"{doc_filename}-picture-{picture_counter}.png"
+            )
+            with element_image_filename.open("wb") as fp:
+                element.image.pil_image.save(fp, "PNG")
 
     end_time = time.time() - start_time
 

docs/examples/export_multimodal.py

@@ -28,6 +28,7 @@ def main():
     # scale=1 correspond of a standard 72 DPI image
     pipeline_options = PdfPipelineOptions()
     pipeline_options.images_scale = IMAGE_RESOLUTION_SCALE
+    pipeline_options.generate_page_images = True
 
     doc_converter = DocumentConverter(
         format_options={