What does binarize mode do?

It runs Otsu thresholding on each page: the brightness histogram is analyzed to find the split that best separates ink from paper, then every pixel becomes pure black or pure white. Gray paper, shadows, and yellowing disappear.

When should I use grayscale mode instead?

Use it for pages with photos, diagrams, or pencil shading you want to keep. It stretches the contrast histogram so paper lifts toward white and ink sinks toward black, without collapsing everything to two colors.

What is stroke bolding?

Morphological dilation — it expands dark character edges by up to 3 pixels. Faint pen strokes and thin fonts that would vanish on an e-ink screen come out readable. It only applies in binarize mode.

Does it work on regular (non-scanned) PDFs?

Yes. Vector pages are rasterized at high resolution, run through the same pipeline, and recompiled. The output is an image-based PDF, so text is no longer selectable.

Will the file get bigger?

Usually the opposite. Binarized pages are stored as PNG, which compresses black-and-white content extremely well — a noisy grayscale scan often shrinks dramatically once it's bilevel.

Which e-readers does this help?

Any monochrome e-paper device — Kindle, Kobo, Onyx Boox, reMarkable — plus anyone printing a scan who wants clean black-on-white output that saves toner.

Is my document uploaded?

No. All processing runs locally in your browser; nothing is sent to a server.

インクけPDF｜スキャンをくっきり｜Kindle・Kobo

グレーでれたPDFスキャンを、インクでみやすいページに。Otsuでをくばし、でいやかすれたをく。Kindle・Koboけにでえ、アップロードでブラウザだけでにできます。

1. Upload a PDF

Drop a scanned or gray-background PDF headed for a Kindle, Kobo, or Boox.

2. Tune the pipeline

Pick binarize or grayscale, then balance contrast and stroke bolding on the live preview.

3. Optimize

Pages are reprocessed and recompiled into a crisp, e-ink-ready PDF.

About the e-Ink Optimizer

Why PDFs Look Bad on e-Paper

E-ink screens render a handful of gray levels and refresh slowly. A scanned page with a cream or gray paper background becomes a muddy rectangle; faint pencil or thin pen strokes drop below the threshold the display can show; and JPEG-style compression noise turns into visible dithering. The file that looked fine on a monitor is barely readable on a Kindle.

The fix is to decide, per pixel, whether it's ink or paper — and that's what this tool does. Otsu thresholding analyzes each page's brightness histogram and picks the split that best separates text from background, then bleaches everything above it to pure white and pushes everything below it to black.

The Optimization Pipeline

Four stages, each tuned for monochrome e-paper.

Otsu Binarization

Histogram analysis finds the ink/paper split automatically — no fixed threshold that fails on dark or yellowed scans.

Stroke Bolding

Morphological dilation expands character edges up to 3 px, so faint handwriting and thin fonts survive on low-contrast screens.

Contrast Control

One slider balances background removal against ink strength, with a live before/after preview on every page.

Smaller Files

Bilevel pages compress far better than noisy scans — the optimized PDF is usually smaller than the original.

Binarize or Grayscale

Binarize mode is the aggressive option: pure black text on pure white, the cleanest possible result for text-heavy documents and the smallest output. Grayscale mode keeps shading — photos, diagrams, pencil texture — but stretches the histogram so the paper tint lifts toward white and ink sinks toward black.

Vector PDFs work too. Pages are rasterized at high resolution in the background, processed, and recompiled, so a born-digital document gets the same e-ink treatment as a scan.

How It Compares

Grayscale conversion and brightness sliders can't isolate ink from paper the way histogram thresholding does.

機能	e-Ink Optimizer	Grayscale conversion
Separates ink from paper tint
Pure black-on-white output
Bolds faint strokes
Keeps photos readable	Grayscale mode
Shrinks file size		Sometimes
No file upload required

Your Documents Stay on Your Device

Manuscripts, exam sheets, and scanned contracts are exactly the files you don't want on a stranger's server. The whole pipeline — histogram analysis, thresholding, dilation, recompression — runs in your browser, and closing the tab wipes everything.