How to Extract Text From Scanned PDFs and Documents
Summary
- Scanned PDFs are images, so you need OCR (optical character recognition) to extract selectable, searchable text.
- You can extract text using built-in tools (depending on your device/app) or third-party OCR apps and online services.
- Results depend heavily on scan quality, language, fonts, layout complexity, and whether the scan is skewed or low-contrast.
- For tables and forms, you may need a workflow that exports to Word/Excel or preserves layout, not just plain text.
- After OCR, save the extracted text somewhere searchable so you can reuse it without re-running OCR each time.
To extract text from a scanned PDF or scanned document, run OCR to convert the image of text into real, selectable text you can copy, search, and paste.
Why you cannot copy text from a scanned PDF (until you OCR it)
A scanned PDF is frequently just a set of images (like photos of pages) stored inside a PDF container. When you drag your cursor over the page and nothing highlights, that is the giveaway: there is no text layer to copy.
OCR (optical character recognition) analyzes the pixels in the scan, detects characters and words, and outputs text. Some tools create:
- A searchable PDF (original image stays, plus an invisible text layer you can select/search).
- Editable text (export to TXT, Word, Google Docs, etc.).
- Structured output (attempts to preserve columns/tables, sometimes exporting to Excel).
Quick check: is your PDF scanned or already text-based?
- Try selecting a line: If you can highlight words cleanly, it is likely text-based (no OCR needed).
- Try searching (Ctrl+F / Cmd+F): If search finds nothing you can see on the page, it may be scanned.
- Zoom in: If letters look slightly blurry or pixelated like a photo, it is likely a scan.
If it is already text-based, you can usually copy/paste directly. If it is scanned, continue with OCR steps below.
Method 1: Use built-in OCR on your device (fastest when available)
Many modern operating systems and document apps include some form of text recognition, but the exact steps vary by version and app. Look for options like “Recognize text,” “Scan to text,” “Copy text from image,” “Make searchable,” or “Export as editable.”
On a phone (iPhone/Android)
- If you only have paper: Use your camera or a scanning app to capture a clean scan (good lighting, flat page, no shadows).
- Extract text: Many camera/gallery or scanning apps can detect text in images and let you copy it. Some can also OCR a PDF after scanning.
Best for: short documents, receipts, letters, quick copy/paste.
On Windows or macOS
- Look for OCR in your PDF viewer: Some viewers can “recognize text” or “make searchable.”
- Look for OCR in built-in capture tools: Some screenshot or image viewers can copy text from an image region.
Best for: occasional OCR when you do not need advanced layout preservation.
Method 2: Use a dedicated OCR tool (best for accuracy and exports)
If you need reliable results, multiple languages, or exports that preserve formatting, a dedicated OCR product or service can be a better fit. When choosing one, focus on what you actually need to output:
- Plain text extraction: fastest, easiest to clean up.
- Editable document export: useful for rewriting or reformatting (Word/Docs).
- Table extraction: useful for invoices, statements, spreadsheets (Excel/CSV).
- Batch processing: useful for many PDFs at once.
Privacy note: Some OCR tools run locally; others upload files to a server for processing. If your document contains sensitive information, check where processing happens and what gets stored.
Method 3: Use an online OCR website (convenient, but be careful)
Online OCR sites can be convenient for one-off tasks, especially if you cannot install software. The tradeoff is that you are uploading your document to a third party.
Use this approach when: the document is not sensitive and you need a quick conversion.
Avoid this approach when: the document includes personal identifiers, financial data, medical information, or confidential business content.
Step-by-step: a reliable OCR workflow that works with any tool
Even though buttons and menus differ, the workflow is consistent. Use this as your checklist.
Step 1: Start with the best scan you can
- Resolution: Higher resolution scans generally OCR better than low-resolution scans.
- Straighten and crop: Skewed pages reduce accuracy; crop out borders and background.
- Improve contrast: Faint text, shadows, and colored paper can confuse OCR.
- Remove motion blur: If you photographed the page, retake it if letters look smeared.
Step 2: Choose the right OCR language(s)
If the tool lets you select language, do it. OCR accuracy drops when the language model does not match the document (for example, English vs. French, or mixed-language documents).
Step 3: Pick your output format based on your goal
- Goal: copy a paragraph into an email → export/copy as plain text.
- Goal: edit the document → export to an editable document format (layout may still need cleanup).
- Goal: reuse the PDF but make it searchable → create a searchable PDF (image + text layer).
- Goal: extract a table → use a table-aware export if available; otherwise expect manual fixes.
Step 4: Proofread and clean up common OCR errors
Plan to review the output. Common issues include:
- Confused characters: O/0, I/1, rn/m, S/5.
- Hyphenation and line breaks: words split across lines or pages.
- Columns: text read left-to-right across columns instead of top-to-bottom per column.
- Headers/footers: repeated page elements mixed into the body text.
Step 5: Save the extracted text somewhere you can find again
OCR is only half the job. If you will need the same extracted text later (for a report, a claim, a case file, a proposal, or repeated form-filling), store it in a searchable place so you do not have to re-run OCR and re-clean the output.
Practical examples: extracting text from common scanned documents
Example A: Copy a clause from a scanned contract PDF
- Run OCR to produce selectable text (searchable PDF or plain text export).
- Search within the OCR output for a distinctive phrase (party name, section title).
- Copy the clause and paste it into your document.
- Manually verify against the scan to ensure no OCR mistakes changed meaning (numbers, dates, “not,” “shall,” etc.).
Example B: Extract line items from a scanned invoice
- Run OCR with a table-aware export if your tool supports it.
- If the table collapses into a paragraph, export as plain text and reformat manually, or try a different OCR tool/output option.
- Double-check totals, tax, and invoice numbers; these are frequent OCR failure points.
Example C: Copy text from a scanned handwritten note
Handwriting recognition is more variable than printed OCR. If your tool supports handwriting OCR, results can be usable for neat handwriting, but you should expect more manual correction. If accuracy is poor, consider transcribing key parts manually.
Limitations to expect (so you can choose the right approach)
- Low-quality scans: faint, noisy, or compressed scans reduce accuracy.
- Complex layouts: multi-column newsletters, forms, and footnotes can be misread in order.
- Non-Latin scripts and mixed languages: require correct language selection and a capable OCR engine.
- Tables: extracting clean rows/columns is harder than extracting plain text.
- Security restrictions: some PDFs are locked or protected; OCR may require a different workflow or permissions.
Where CopyCharm fits: capture OCR output and reuse it without redoing the work
Once you have extracted text from a scanned PDF (by any OCR method), the next friction point is reuse: you may need the same paragraph, address block, policy language, or cleaned-up table text multiple times across emails, documents, and AI chats.
CopyCharm is a Windows desktop app that saves copied text locally so you can search past clips, favorite important clips, and separately save reusable prompts. A practical OCR-to-reuse workflow looks like this:
- Save: After OCR, copy the cleaned text (or key excerpts) and copy it to your clipboard. CopyCharm can keep that copied text so it is available later.
- Find: When you need it again, search your past clips in CopyCharm and open the exact excerpt you previously copied.
- Reuse: Copy/paste the excerpt into your destination (email, Word, a form field, a ticket, or a chat).
If you use ChatGPT and want it to reference your extracted text, CopyCharm also has an authenticated ChatGPT connector backed by optional AI Access sync. After you sign in with an eligible active CopyCharm purchase, authorize the CopyCharm Desktop connection, enable and complete AI Access sync, and authorize the ChatGPT connector, ChatGPT can search and retrieve only supported Synced Data (Favorite Clips, Saved Prompts, and optional Other Clips within your selected time range). ChatGPT cannot access unsynced local CopyCharm data, and retrieval is user-directed.
For Claude, Gemini, Cursor, email, documents, and other applications, the workflow is manual: search or retrieve the text in CopyCharm, then copy/paste it into the destination.
Try CopyCharm for saving and reusing extracted text
Frequently Asked Questions
FAQ 1: Why can’t I copy text from my scanned PDF?
Answer: Because the “text” is actually an image of a page. There is no text layer for your computer to select. You need OCR to convert the image into real text you can highlight, search, and copy.
Takeaway: If selection doesn’t work, run OCR first.
FAQ 2: What’s the difference between a searchable PDF and an editable document?
Answer: A searchable PDF usually keeps the original scanned image and adds an (often invisible) text layer so you can search and copy text while preserving the look of the scan. An editable document export (like a word-processing file) tries to rebuild paragraphs, spacing, and layout so you can edit it, but it may require cleanup.
Takeaway: Choose searchable PDF for archiving/search; choose editable export for rewriting.
FAQ 3: How can I improve OCR accuracy on a blurry or skewed scan?
Answer: If you can rescan, prioritize a straight, well-lit capture with higher resolution and good contrast. If you can’t rescan, try deskew/straighten, crop margins, and adjust contrast before OCR. Also select the correct OCR language(s) if your tool offers that option.
Takeaway: Better input quality and correct language selection usually matter more than tool-hopping.
FAQ 4: Can OCR extract tables from scanned PDFs into Excel?
Answer: Sometimes. Table extraction depends on scan clarity and whether the OCR tool supports table-aware output. Even with the right tool, you may need to fix merged cells, misread numbers, or column alignment after export.
Takeaway: Expect to review and correct table exports, especially for invoices and statements.
FAQ 5: Does OCR work on handwriting?
Answer: Handwriting recognition is more variable than printed OCR. Neat handwriting can convert reasonably, while cursive, faint pencil, or messy notes may produce errors. If accuracy is critical, consider manual transcription of key fields (names, dates, amounts).
Takeaway: Handwriting OCR can help, but plan for manual verification.
FAQ 6: Is it safe to use online OCR websites for sensitive documents?
Answer: It depends on the site’s handling of uploads, storage, and retention, which varies by provider. If the document contains sensitive personal or business information, consider an OCR option that processes locally on your device, or confirm the service’s privacy and deletion behavior before uploading.
Takeaway: Treat online OCR as a file upload to a third party and decide accordingly.
FAQ 7: What should I do if OCR output has weird characters or broken line breaks?
Answer: First, compare the OCR text against the scan around problem areas (numbers, names, and headings). Then clean up common issues: remove hyphenation at line ends, join wrapped lines into paragraphs, and correct character confusions like O/0 or I/1. If the document has columns, try an OCR mode/output that better preserves reading order, or extract one column/region at a time if your tool supports region selection.
Takeaway: A quick cleanup pass is normal, especially for columns and low-contrast scans.
FAQ 8: How can CopyCharm help after I’ve extracted text from a scanned PDF?
Answer: After OCR, you can copy the cleaned text and let CopyCharm save that copied text locally so you can search it later, favorite important excerpts, and separately save reusable prompts. If you want ChatGPT to reference specific extracted text, CopyCharm’s authenticated ChatGPT connector can search and retrieve only supported Synced Data after eligible authorization and AI Access sync; it cannot access unsynced local CopyCharm data. For other apps (email, documents, Claude, Gemini, Cursor), you’d retrieve the text in CopyCharm and copy/paste it where you need it.
Takeaway: OCR creates the text; a searchable clip workflow helps you reuse it without repeating OCR and cleanup.