How to remove dark scanner borders from a scanned pdf
- Step 1Measure the border thickness in points — Open the scan and estimate the dark band's width on each edge. Typical flatbed borders are 10–30 pt (
72 pt= 1 inch). Sheet-feed scanners often leave a thinner, more uniform edge. If the band is wider on one side (the hinge side of a flatbed is common), note that side separately. - Step 2Open the crop tool and drop the scanned PDF — Load the file into the PDF Crop tool. Parsing happens locally — the scan is never uploaded. Free accounts handle PDFs up to 2 MB and 50 pages; a multi-page scan can hit that quickly, so compress or upgrade if needed.
- Step 3Enter equal values for a uniform border — If the dark band is the same width all around, type the same number in
top,right,bottom, andleft— start around 15–20 pt for a typical flatbed and adjust. - Step 4Use different values for an uneven border — If one edge is darker or wider (the lid-hinge side, or where a smaller document sat against the platen edge), set that side higher than the others. The four inputs are fully independent.
- Step 5Run the crop and check the worst page — Apply the crop and open the result. Because the same CropBox covers every page, check the page where content sits closest to the edge — if the crop ate into the text or a stamp, you over-cropped.
- Step 6Re-run on the original if needed — The original is never modified, so if the border isn't fully gone (or content got clipped) just run the tool again on the source file with adjusted values.
Typical scanner border widths
Rough starting points — measure your own scan, since border width depends on the scanner, the lid seal, and the document size relative to the platen.
| Scanner / situation | Typical border | Points to try | Notes |
|---|---|---|---|
| Flatbed, full-page document | Thin even shadow | 10–15 pt all sides | Border is mostly the platen edge |
| Flatbed, small document on glass | Wide dark area beyond the doc | Measure per side, often 30–60 pt | Black where the lid covered empty glass |
| Sheet-feed (ADF) | Thin, very uniform | 8–12 pt all sides | Consistent page to page |
| Lid-hinge side darker | Wider band on one edge | Higher value that side only | Lid doesn't seal flat near the hinge |
| Book/spine scan | Dark gutter on inner edge | Higher inner-edge value | Curvature may remain; crop only trims the straight edge |
Crop vs. the right tool for the job
Cropping handles a straight dark border. Other scan problems need other tools.
| Scan problem | Does cropping fix it? | Use instead |
|---|---|---|
| Straight dark border / shadow band | Yes — set a CropBox inside the band | This tool |
| Whole scan is rotated 90°/180° | No | PDF Rotate |
| Scan is huge (many MB) | No — crop doesn't shrink bytes | Lossy compress |
| Need the trim to be permanent | Partly — CropBox is reversible | Flatten after cropping |
| Want selectable text from the scan | No | PDF OCR |
Cookbook
Border-removal settings for the common scanner situations. Values in points; the same crop applies to every page.
Uniform flatbed shadow, all sides equal
A clean full-page flatbed scan with a thin even shadow. Pull each edge in by 15 pt.
Inputs: top: 15 right: 15 bottom: 15 left: 15 A4 scan (595 × 842 pt) becomes: CropBox = (15, 15, 565, 812) → thin dark border hidden on all four edges
Small document on a large platen
You scanned a half-sheet on a full-size flatbed, so the lid left a wide black area on two sides. Crop those edges hard.
Inputs: top: 18 right: 18 bottom: 55 left: 18 → the bottom held the empty-glass black area; cropping it 55 pt removes the wide band there
Lid-hinge side darker than the rest
The hinge edge didn't seal flat, leaving a wider shadow on one side. Raise just that value.
Inputs: top: 14 right: 14 bottom: 14 left: 40 → left edge (hinge side) trimmed extra; other three stay light
Sheet-feed scan, thin uniform edge
An ADF scan with a thin, consistent border across all pages — ideal for one uniform crop.
Inputs: top: 10 right: 10 bottom: 10 left: 10 → same CropBox on every page; consistent result because the feeder produces a consistent border
Crop then make it permanent
For a scan you'll distribute, crop to hide the border, then flatten so a recipient can't reveal it by removing the CropBox.
1. /pdf-tools/pdf-crop → CropBox hides the border 2. /pdf-tools/pdf-flatten → bakes the visible area in → note: flattening rasterises; re-run OCR afterward if you need selectable text
Edge cases and what actually happens
The border is part of the page image, so cropping only hides it
PreservedA scanned page is one big image; the dark band is pixels inside that image. The CropBox hides those pixels from view but the image data is unchanged underneath. A recipient could remove the CropBox and see the border again. To make removal permanent, flatten after cropping; note that flattening rasterises the page anew.
Scan is skewed, so the border is tilted
limitedA rectangular CropBox has straight horizontal and vertical edges. If the page was fed in crooked, the dark border runs diagonally and a straight crop either leaves part of the band or eats into content. Crop can't follow a tilted edge — de-skew the scan in a dedicated scanner/image tool first, then crop the now-straight border.
Border width varies page to page
By designFlatbed scans done one page at a time can have slightly different border widths. The tool applies one CropBox to all pages, so it works best when borders are consistent (typical of an ADF). For inconsistent flatbed scans, pick a value that clears the widest border without clipping content on the tightest page — or split the file and crop the outliers separately.
Cropping into a stamp, signature, or edge note
ClippedScans of forms often have stamps, initials, or notes right at the margin. A border crop that's too aggressive hides them. Because cropping only hides (nothing is deleted), re-run on the original with smaller values, or use an uneven crop that spares the edge holding the content.
Multi-page scan over the 2 MB free limit
Free tier limitImage-based scans are large — a 20-page colour scan easily passes 2 MB, the free-tier cap (also 50 pages). Either upgrade to Pro (50 MB / 500 pages) or shrink the scan first with lossy compression, which is built for image-heavy PDFs, then crop the smaller file.
Scan also has an OCR text layer
SupportedIf your scan carries an invisible OCR text layer, cropping the CropBox doesn't remove that layer — it stays attached to the page. Text behind the cropped-away border simply falls outside the visible area. If you later flatten, the page is rasterised and the text layer is lost; re-run OCR afterward.
left + right (or top + bottom) too large
invalid crop boxThe visible width is width − left − right. On a wide-bordered small-document scan it's tempting to enter big numbers, but if the horizontal pair exceeds the page width the CropBox is zero or negative and viewers may show a collapsed page. Keep each pair below the page dimension.
Scan is rotated 90 degrees
use rotateIf the whole page came in sideways, what looks like the 'top' border is actually a side. Cropping won't fix orientation. Straighten it with PDF Rotate first, then crop the border on the correctly-oriented page.
Frequently asked questions
What crop value removes a typical scanner border?
Most flatbed borders are 10–30 pt wide (72 pt = 1 inch). Try 15 pt on each side for a clean full-page flatbed scan, or 10 pt for a thin ADF border, then adjust. If you scanned a small document on a large platen, the black area can be much wider — measure that edge specifically and crop it harder.
My scanner border is uneven — can I set different values per side?
Yes. The four inputs (top, right, bottom, left) are independent. The lid-hinge edge of a flatbed often has a wider shadow, so raise just that value while keeping the other three low. Likewise, if a small document sat against one edge of the glass, crop the empty-glass side harder.
Does cropping actually delete the dark border pixels?
No — it hides them. Cropping sets a CropBox, the rectangle a viewer displays; the scanned image (border included) is still in the file underneath. A recipient could remove the CropBox and see the border again. To make the trim permanent, flatten the cropped PDF, which rasterises the visible area.
Can I crop the border on every page at once?
Yes — that's the default behaviour. The tool applies one CropBox to all pages, which suits scans because the border tends to repeat consistently page to page (especially from a sheet feeder). There is no per-page control, so if border widths vary a lot, pick a value that clears the widest one.
My scan is crooked and the border is tilted — will cropping help?
Not well. A CropBox is a straight rectangle, so it can't follow a diagonal border on a skewed scan. De-skew the scan in your scanner software or an image editor first to make the border straight, then use this tool to crop it off cleanly.
Will cropping reduce the file size of my scan?
No. The border pixels stay in the file (just hidden), so the byte size barely changes. Scans are large because they're images — to actually shrink one, use lossy compression, which re-encodes each page image. You can crop first, then compress, or vice versa.
Are my scanned documents uploaded anywhere?
No. The PDF is parsed and cropped entirely in your browser — nothing is sent to a server. That's important for the documents people scan: IDs, bank statements, medical records, signed forms. Only an anonymous usage counter is recorded when you're signed in, never the file.
Will the cropped scan print without the black border?
Yes — printers print the CropBox area, so the dark band falls outside it and isn't laid down on the page. That's the main practical win: no wasted toner on a black border every page. Just don't crop so tight that content near the edge gets clipped on print.
How do I keep an OCR text layer when removing the border?
Cropping alone keeps the text layer — it only changes what's displayed, not the page's text content. The risk is later flattening: PDF Flatten rasterises the page and drops the invisible text. If you flatten, re-run OCR afterward to restore selectable text.
How big a scanned PDF can I process?
Free accounts allow up to 2 MB and 50 pages — easy to exceed with a colour multi-page scan. Pro raises it to 50 MB / 500 pages. If a scan is over the limit only because of size, compress it with lossy compression before cropping.
What if I crop too much and hide part of the document?
Nothing is lost — cropping only hides content outside the CropBox, and the original file is never modified. Just run the tool again on the source scan with smaller values. Leave a few points of buffer between the crop line and where content begins.
Can I crop a book or spine scan with a dark gutter?
You can crop the dark gutter on the straight inner edge by raising that side's value. What cropping can't fix is page curvature near the spine, which often comes with book scans — that distortion lives in the image and needs a dedicated book-scan correction tool. Crop trims the straight border; it can't flatten a curved page.
Privacy first
All PDF processing runs locally in your browser using PDF-lib and pdf.js. No file is ever uploaded — only metadata counters are saved for signed-in dashboard stats.