OCR & Accessibility

Make your scanned documents searchable and accessible to all users using OCR (Optical Character Recognition) and accessibility features.

Understanding OCR

OCR (Optical Character Recognition) is a technology that recognizes text within images. In the PDF Editor, OCR is used to:

• Extract text from scanned pages — Convert image-based PDFs to searchable documents
• Make text selectable — Allows viewers to copy and select text
• Enable searching — Find specific words or phrases within the document
• Preserve original image — The scanned image remains visible while text becomes available underneath

When to Use OCR

Use OCR when:

• Scanning documents — You have scanned paper documents as PDFs
• Faxed documents — Converting old faxes to text
• Screenshots with text — Extracting text from screenshot-based PDFs
• Archived documents — Digitizing older documents for searchability

Don't need OCR for:

• Documents created in Word, Excel, or other applications
• PDFs exported from software (already have embedded text)
• Digital forms or documents

Running OCR on a Page

Starting OCR

1. Navigate to the page you want to make searchable
2. Look for the "Make searchable (OCR)" button above the page thumbnail
3. Click the button
4. An OCR process starts — this may take a few seconds

During OCR

• Processing indicator appears while the text is being recognized
• First run warning — First use may take longer as Tesseract.js downloads
• No interruptions — You can continue editing other elements while OCR runs

After OCR Completes

1. The recognized text appears in a new text layer
2. OCR text is dimmed and outlined in the editor for visibility
3. Text is marked as OCR so it exports correctly
4. The text becomes selectable and searchable in any PDF viewer

How OCR Works

Recognition Process

1. The PDF Editor renders the page as an image
2. Tesseract.js analyzes the image to find text
3. Each line of recognized text becomes an element
4. Text is positioned to match its location on the page
5. Text opacity is set to 0% so it's invisible but selectable in export

Accuracy

OCR accuracy depends on:

• Image quality — Clearer scans produce more accurate results
• Text size — Larger text is recognized more accurately
• Font type — Standard fonts are recognized better than decorative fonts
• Language — Currently optimized for English

Typical accuracy:

• Good scans — 95%+ accuracy
• Average scans — 85-95% accuracy
• Poor scans — 60-85% accuracy

Editing OCR Results

Reviewing OCR Text

1. After OCR completes, review each recognized line
2. OCR text appears dimmed with an outline
3. Look for any misrecognitions or errors

Fixing Errors

1. Double-click any OCR text element to edit it
2. Correct any misrecognized characters
3. Click outside to save the changes
4. The corrected text will be searchable in the exported PDF

Removing Incorrect Results

1. Click an OCR text element to select it
2. Press Delete to remove it
3. Optionally add correct text using the Text tool

Adding Missing Text

If OCR misses some text:

1. Use the Text tool to add missing content
2. Position it in the same location as the OCR'd text
3. Use the same font size and style for consistency

OCR Text in Exports

How OCR Text is Exported

When you export your document:

1. OCR text is included in the PDF with 0% opacity
2. The text is invisible to viewers
3. The text is fully selectable and searchable
4. The original scanned image remains visible
5. Users can copy OCR'd text to their clipboard

Searching OCR'd Documents

In any PDF viewer, users can:

1. Ctrl+F (or Cmd+F) to open the search dialog
2. Search for any recognized text
3. The text is found and highlighted
4. Original image remains visible underneath

This makes scanned documents as searchable as digitally created ones.

Batch OCR

Currently, OCR runs on one page at a time. For multiple pages:

1. Navigate to each page
2. Click the OCR button for each page
3. Wait for completion before moving to the next page

Tip — Start with important pages first, then continue with others.

Performance Considerations

Download Size

First-time OCR requires downloading Tesseract.js:

• Initial download — ~4-6 MB (one-time, cached by browser)
• Subsequent runs — Use cached version (very fast)

Processing Time

• First run — 5-15 seconds per page
• Subsequent runs — 2-5 seconds per page
• Large documents — Consider processing during off-peak hours

Browser Resources

• CPU usage — Processing uses your computer's CPU
• Memory usage — Minimal; processing runs in a web worker
• Battery — May impact laptop battery life during processing

Accessibility Features

Beyond OCR, the PDF Editor supports several accessibility features.

Text and Visual Hierarchy

1. Use adequate font sizes — Minimum 12pt for body text
2. Clear headings — Use the Text tool to create distinct heading levels
3. High contrast — Ensure text color contrasts well with background
4. Consistent styling — Use uniform fonts and sizes for similar content

Reading Order

For documents with complex layouts:

1. Arrange elements logically — Top to bottom, left to right
2. Use spacing — Group related content together
3. Number pages — Add page numbers for reference

Semantic Meaning

1. Describe images — Add nearby text explaining what images show
2. Use lists — Structure information clearly
3. Provide context — Ensure text is understandable without visual elements

Color and Contrast

• Don't rely on color alone — Use color + text or patterns
• Check contrast — Use a contrast checker tool for critical documents
• Light backgrounds — Help readers with low vision

For Scanned Documents

After running OCR:

1. Review results — Ensure text recognition is accurate
2. Correct errors — Fix any misrecognitions
3. Maintain structure — Don't rearrange elements after OCR

PDF Accessibility Standards

The PDF Editor helps create accessible documents that meet:

• WCAG 2.1 — Web Content Accessibility Guidelines
• Section 508 — US accessibility standards for digital content
• PDF/A compliance — Archival format standard

Tips for Accessible Documents

Text-First Approach

1. Add text rather than relying on images or shapes
2. Use proper text formatting instead of visual effects
3. Structure with text hierarchy for clarity

Image Handling

1. Describe all images with nearby text
2. Use alt text in text elements near images
3. Avoid text in images — text in images can't be OCR'd

Form Fields

1. Add text labels for all form fields
2. Space fields appropriately for readability
3. Use contrasting background colors for field areas

Testing Accessibility

1. Read content aloud — Use a screen reader to check flow
2. Test with zoom — Ensure document is readable at 200% zoom
3. Check color — Verify usability in grayscale

Troubleshooting OCR

OCR doesn't seem to work

• Why — Text may already be present in the PDF
• Solution — OCR is only needed for scanned/image-based pages
• Check — Try searching for text first; if found, OCR isn't needed

OCR results are very inaccurate

• Why — The scan quality is low, or text is too small
• Solution — Manually add correct text using the Text tool
• Prevention — Use higher quality scans in the future

OCR is very slow

• Why — First run requires downloading Tesseract.js
• Solution — Wait for completion; subsequent runs are faster
• Optimization — Close other browser tabs to free resources

OCR creates duplicate text

• Why — Both original text and OCR'd text might be present
• Solution — Delete the OCR text layer if original is already searchable
• Avoid — Run OCR only once per page

Browser Requirements

OCR requires:

• Modern browser — Chrome, Firefox, Safari, or Edge (recent versions)
• JavaScript enabled — Required for OCR processing
• Sufficient memory — At least 2 GB RAM recommended
• Internet connection — First run downloads Tesseract.js

Keyboard Shortcuts

Shortcut	Action
Ctrl+F / Cmd+F	Search in exported PDF
Ctrl+Z / Cmd+Z	Undo OCR text additions

Next Steps

• Sharing & Collaboration — Share accessible documents
• Text Editing — Format text for accessibility
• FAQ & Troubleshooting — Get more help