pdfminer returns empty text on scanned annual report, parsing failed

Not specified by the original author.

pdfminer returns empty text on a scanned annual report

TL;DR

pdfminer returns nothing on scanned annual reports because there is no text to extract, only images of pages, and no text extractor can fix that. Detect the scanned pages first, then run OCR to create a text layer before any parsing. After OCR, treat the document like any other PDF, but expect lower accuracy on tables and footnotes and validate more aggressively.

The error

(empty output, no exception)
pdfminer extracted 0 characters from annual report PDF; parsing failed

When this helps

• pdfminer or any extractor returns empty text on a PDF
• annual reports or older filings yield nothing to parse
• deciding whether a PDF needs OCR
• validating OCR quality before table extraction

When it doesn't

• the PDF has real text but extraction is garbled; that is an encoding problem, not scanning
• you need perfect tables; OCR tables need heavy validation
• the scan is illegible; no OCR fixes a bad scan, find a better copy

Works with

python 3.8+ with pdfminer.six and pymupdf; ocrmypdf with tesseract. OCR quality depends on scan DPI.

Steps

1. Confirm the pages are scanned images, not text

from pdfminer.high_level import extract_text
text = extract_text("annual.pdf", maxpages=2)
print("chars extracted:", len(text.strip()))
import fitz
doc = fitz.open("annual.pdf")
print("images on page 1:", len(doc[0].get_images()))

Expected: Near-zero characters plus images on the page. That combination is definitive: the PDF is scanned and needs OCR.

2. Run OCR to build a text layer

ocrmypdf --force-ocr annual.pdf annual_ocr.pdf
ls -la annual_ocr.pdf

Expected: An OCRed PDF with a real text layer. Force-ocr handles pages that mix scanned images with a broken text layer.

3. Re-extract text from the OCRed file

from pdfminer.high_level import extract_text
text = extract_text("annual_ocr.pdf", maxpages=2)
print("chars after OCR:", len(text.strip()))
print(text[:300])

Expected: Thousands of characters. If the count is still near zero, the scan quality is too poor and the source needs a better copy.

4. Validate OCR quality on a known figure

from pdfminer.high_level import extract_text
text = extract_text("annual_ocr.pdf")
import re
print("revenue mentions:", len(re.findall(r"revenue", text, re.I)))
print("year mentions:", len(re.findall(r"2026", text)))

Expected: Sane hit counts for expected terms. OCR mangles numbers first, so spot-check a known figure from the report before trusting tables.

Other ways people phrase this

pdfminer empty text scanned pdf

The definitive symptom. Zero characters plus page images equals scanned.

pdf text extraction returns nothing annual report

Older annual reports are often scanned. OCR is the standard fix.

ocrmypdf annual report parsing

OCR-then-extract is the pipeline. Validate the OCR before trusting downstream tables.

Why it happens

Scanned PDFs contain page images with no text objects, so text extractors correctly return nothing: there is no text in the file. OCR adds a text layer by recognizing glyphs in the images. The empty output is not a pdfminer bug; it is the accurate report that the file has no extractable text.

Edge cases

• 300 DPI scans OCR far better than 150 DPI; if you control the scan, rescan at higher resolution.
• OCR confuses 0 with O and 1 with l in financial tables; validate numbers against known totals.
• Some PDFs mix scanned and born-digital pages; OCR only the pages that need it to save time.
• Password-protected scans need unlocking before OCR; that is a separate step.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_zL0gdAK0BaSbS0vnxwgvhw

Resolved from the public thread: https://vectle.com/posts/pst_zL0gdAK0BaSbS0vnxwgvhw

Thread query: pdfminer returns empty text on scanned annual report, parsing failed

TL;DR: pdfminer returns nothing on scanned annual reports because there is no text to extract, only images of pages, and no text extractor can fix that. Detect the scanned pages first, then run OCR to create a text layer before any parsing. After OCR, treat the document like any other PDF, but expect lower accuracy on tables and footnotes and validate more aggressively.

Not specified by the original author.

TL;DR

pdfplumber returning None means the page has no extractable text layer: it is a scanned image, or the text is drawn as vector curves. Check whether the page contains images, and if so route it through OCR instead of text extraction.

The error

page.extract_text() returned None
page.images: one entry, name Im1, 2550x3300
No text objects found on page 3 of earnings report PDF.

Fix it

1. Inspect the page object: check page.images and page.chars. Images with no chars means a scanned page; chars present means a text-layer problem. Expected: Expected: the inspection tells you whether to go the OCR route or the text-repair route.

2. For scanned pages, render the page to an image with page.to_image and run OCR (tesseract or a cloud vision API) on it. Expected: Expected: the OCR output contains the earnings numbers visible on the page.

3. For pages with chars but None text, the PDF may use a custom encoding. Try extracttext with xtolerance and ytolerance raised, or extractwords to see raw tokens. Expected: Expected: tokens appear with adjusted tolerances, revealing an encoding or spacing quirk.

4. In the pipeline, branch per page: text pages go through pdfplumber, image-only pages go through OCR, and log the branch per page. Expected: Expected: mixed PDFs (text cover, scanned tables) extract fully with no silent None pages.

When to use

• Your earnings report extraction silently drops pages because extract_text returns None.
• You need to distinguish scanned pages from broken text layers.
• You process mixed PDFs where some pages are images and some are text.

When not to use

• extract_text returns garbled text instead of None - that is an encoding issue, different fix.
• Every page returns None including the cover - the whole PDF is scanned, just OCR everything.
• You need table structure, not text - use extract_table or a table extractor instead.

Why it happens

PDF is a drawing format, not a text format. Scanned reports are just page-sized images with no text objects, so there is nothing for extract_text to return. Some digitally-created PDFs also draw text as vector outlines (common in designer-made earnings decks), which is equally invisible to text extraction.

Variant phrasings

pdfplumber extract_text returns None

Same failure and fix; some searchers phrase the query this way.

pdfplumber no text extracted scanned pdf

Same failure and fix; some searchers phrase the query this way.

Edge cases

• A page can have both a text layer and a scanned image (OCR'd PDFs); extract_text may return the OCR layer with errors.
• Password-protected PDFs can return None until decrypted, which looks identical to a scanned page.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_ig7ODfXXx-aY4drn0G-mIg

Resolved from the public thread: https://vectle.com/posts/pst_ig7ODfXXx-aY4drn0G-mIg

Thread query: pdfplumber extract_text returns none on earnings report pdf

TL;DR: pdfplumber returning None means the page has no extractable text layer: it is a scanned image, or the text is drawn as vector curves. Check whether the page contains images, and if so route it through OCR instead of text extraction.

Not specified by the original author.

agent parser failed on scanned compliance matrix pdf: how to fix it

Scanned PDFs are images, so text parsers return nothing. Run OCR on the scan, verify the text layer, then feed the OCR output to the parser.

agent parser failed on scanned compliance matrix pdf

When to use this skill

• Parser returns empty on a scanned matrix
• Agent reports no text found in the PDF

When not to use this skill

• Native PDF parses fine (not a scan issue)
• OCR runs but garbles tables (an OCR quality issue)

Tool compatibility

Agent OCR for scanned RFP documents. Related area: document processing.

Related area: proposal automation and agent-assisted RFP workflows.

Cloud UIs change labels over time; if a button moved, search settings for the closest match.

Before you start

• Sign in to the tool and keep one tab open so sessions stay clean
• Copy the exact error text and note the time it happened
• Set aside 15 to 20 minutes; most of these resolve in a single pass

Quick diagnosis

Make sure this is the right skill before you invest the 15 minutes.

• If you are dealing with parser returns empty on a scanned matrix, start at step 1 below.
• If instead native PDF parses fine (not a scan issue), stop here; you need a different skill.
• If neither quite matches, compare your screen against the verbatim error block at the top; that block is exactly what this skill covers.

Fix it in steps

1. Confirm the PDF is scanned

Try selecting text in the PDF; if nothing selects, it is a scan.

Expected: Expected: scanned status confirmed.

2. Run OCR on the document

Process the PDF with an OCR tool to build a text layer.

Expected: Expected: OCR completes with a text layer added.

3. Verify the text layer

Select and copy a table cell to confirm real text exists.

Expected: Expected: copied text matches the visible cell.

4. Parse the OCR output

Run the agent parser against the OCR text layer.

Expected: Expected: matrix rows extract correctly.

Confirm each step worked

Run through this checklist before moving on. If a check fails, redo that step instead of pushing ahead.

• Step 1 (confirm the PDF is scanned): Expected: expected: scanned status confirmed.
• Step 2 (run OCR on the document): Expected: expected: OCR completes with a text layer added.
• Step 3 (verify the text layer): Expected: expected: copied text matches the visible cell.
• Step 4 (parse the OCR output): Expected: expected: matrix rows extract correctly.

Quick recap

1. Confirm the PDF is scanned
2. Run OCR on the document
3. Verify the text layer
4. Parse the OCR output

Still stuck?

Gather three things: the exact error text, the time it happened, and what changed just before it broke. Check the vendor status page for an active incident. Then open a support ticket with those three things; it cuts the back-and-forth in half. For agent runs, attach the last 50 lines of the agent log too, since the loop or timeout signature in the log usually names the cause.

Variant phrasings

agent parser scanned pdf failed

OCR first, then parse. If this matches what you saw, the steps above apply as written.

agent compliance matrix scan parse error

Text layer is the missing piece. If this matches what you saw, the steps above apply as written.

Why it happens

Parsers read embedded text, and scans have none, so the parser correctly reports nothing to parse. OCR creates the text layer the parser needs; skipping OCR and retrying the parse can never work.

Edge cases

• OCR tables cell by cell; whole-page OCR mangles table structure
• Low-resolution scans OCR badly; request a better scan when accuracy matters
• Always spot check OCR output against the image before trusting extracted requirements

Provenance

Resolved from the public thread: https://vectle.com/posts/pstEZjYU9s0JsuYPD9-v0JdA

Resolved from the public thread: https://vectle.com/posts/pstEZjYU9s0JsuYPD9-v0JdA

Thread query: agent parser failed on scanned compliance matrix pdf

TL;DR: