What this enhancer actually does
Most "enhance scanned pdf" tools online are either fake (they compress the PDF and call it enhancement) or hidden behind a paid Acrobat Pro subscription. This one does real per-page image processing:
- Auto — the recommended default. Builds a luminance histogram of every page, finds the "paper white" peak, and remaps the whole page so paper becomes clean bright white and ink becomes clean dark. Handles mixed colour and grayscale scans well.
- Black & White — computes an adaptive threshold per page (mean-based, Otsu-ish) so every pixel becomes either 0 or 255. Best for pure text scans where you want the smallest possible file and the cleanest look. Bonus: this dramatically improves OCR accuracy if you then run the output through /ocr.
- Grayscale — strips colour cast, boosts luminance contrast ~35%. Good for scans with faint pencil marks, or mixed text-and-photo pages where you don't want the photos threshold-crushed.
- Gentle — mild contrast boost (+15%) with no background cleanup. Use when the scan is already pretty good and you just want to give it a small kick.
When to use each preset
Auto: contracts, reports, book pages, mixed documents — anything with a paper background and some ink. Safe default.
Black & White: handwritten notes, receipts, form fills, textbook pages, faded scans of typed text. Also the preset to use before running the PDF through OCR — most OCR engines are twice as accurate on clean B&W input.
Grayscale: photos scanned in the middle of a document, art scans, blueprints, x-rays, mixed content where killing all colour would kill the useful part of the page.
Gentle: already-clean scans that just look a bit washed out. Also the safest choice if you're unsure — it can't make a good scan worse.
Improving a scanned PDF for OCR
If your goal is to feed the scan into an OCR tool (to make it searchable, extract the text into Word, or copy quotes from a PDF that was originally on paper), the enhancement step matters enormously. OCR engines look for high-contrast text against a clean background. A faded scan with grey background gives ~60% recognition accuracy; the same scan after Black & White enhancement typically hits 90%+.
Recommended workflow: run through this tool with the Black & White preset first, then feed the enhanced PDF into our OCR tool or our PDF to Word converter. Two tools, two clean steps, no upload in either.
Shrinking a huge scanned PDF
Scanned PDFs are often ridiculously large — a 20-page scan can be 100+ MB because it's stored as full-colour, high-DPI bitmap of every page. Running through this enhancer with the Black & White preset routinely shrinks the file 5-10× (a 100 MB scan becomes 10-20 MB) because the JPEG re-encoding at quality 90% + threshold cleanup produces very compressible pages. Handy for emailing large scans or uploading to systems with size limits.
How the processing pipeline works
The tool loads your PDF with pdf.js (the same engine Firefox uses), renders each page onto a browser canvas at 144 DPI so the enhancement has enough detail to work with, extracts the raw RGBA pixel buffer, applies the chosen filter as pure JavaScript operations on the pixel data, re-encodes the result as JPEG at 90% quality, and finally uses pdf-lib to assemble a new PDF with the enhanced pages. Everything happens in your browser — no server, no upload, no third party.
Privacy — your scan stays on your device
Scanned PDFs are often the most sensitive documents people have: bank statements, medical records, contracts, government forms, ID copies. Uploading them to a random "enhance scanned PDF online" site means that site now has a copy — even if they delete it after processing you're trusting their word. This tool runs entirely in your browser. Verify by opening DevTools → Network tab during enhancement; no image or PDF uploads appear.
What it does not do (yet)
Honestly: no deskew (auto-rotation of crooked scans), no deblur, no perspective correction of skewed page shots. Those are on the roadmap. For now, straighten heavily-rotated pages in your PDF viewer first (Preview's Rotate Right, Acrobat's Rotate Pages), then enhance. If you need perspective correction on a photo of a page, use our PDF scanner instead — that tool has full edge detection and perspective warp.