HomePlatformsDocumentsGlossaryBlogContactAbout
✂️
KYC Error

Document Cropped

Understanding and fixing this common KYC rejection

⚠️ What This Error Means

This error means part of your ID physically falls outside the photo frame, usually a corner or one full edge, so the KYC system cannot confirm it is looking at a complete document. It triggers the instant any edge detection algorithm finds the document boundary touching or crossing the image border instead of sitting inside it with margin on all sides.

Verification systems insist on all four corners being visible because that is how they confirm the aspect ratio matches a known passport, license, or national ID template, and how they check the security details printed near the border. Many ID documents carry microprint text, guilloche line patterns, or color bands that run close to the edge, and a document reader uses those border details as part of its authenticity scoring. Cut off a corner and the system loses both the boundary it needs for perspective correction and the fine detail it needs to confirm nothing was altered or lifted from the original.

More often than people expect, this is not a framing mistake at all, it is the phone's own camera software doing it. Scanner-style apps run automatic edge detection to find the document and crop tight to that boundary for a cleaner-looking result. That detection gets confused by glare on a corner, a cluttered or similarly colored background, or a slight tilt in the shot, and it draws the crop line a few millimeters inside the real edge. You never see it happen because the app shows you a nicely cropped rectangle, not a warning that it just guessed wrong about where your ID actually ends.

✅ How to Fix It

  1. Shoot in landscape, not portrait. Landscape orientation matches the proportions of most ID cards and passport photo pages far better than portrait, which gives all four corners room to sit inside the frame instead of crowding the edges.
  2. Leave 5 to 10mm of margin on every side. Do not zoom in or fill the frame edge to edge. A visible strip of background around the entire document gives auto crop room to work without cutting into the actual ID.
  3. Use a plain, contrasting background. Set the document on a solid dark surface if it is light colored, or vice versa. A cluttered or similarly toned background is the single biggest reason edge detection guesses wrong about where the document ends.
  4. Hold the camera flat and directly overhead. Shooting at an angle makes the far corners appear smaller and pushes them closer to the edge of the frame, which is exactly what gets trimmed first.
  5. Check the actual preview before you submit. Zoom into the image after taking it and confirm all four corners are fully visible with clean background around them. If your camera app applied its own auto crop, look closely, since the displayed thumbnail is not always the same file that gets uploaded.

Cropped vs Other Similar Errors

These four rejections get mixed up constantly because they can look similar in a quick glance at your photo. If you landed on this page but something feels off, use this to find the right fix.

This Looks Like Actually Is How to Tell Apart
A corner or edge is missing Document Cropped The document boundary itself runs off the frame. You physically cannot see where the card or page ends on one or more sides, even if the visible part looks sharp.
Whole ID visible but looks soft or pixelated Low Resolution All four corners are present, but zooming in shows blocky pixels or a fuzzy, low-detail image everywhere, not just at the edges. This is a camera or upload quality issue, not a framing issue.
System says it cannot read the text or data Unable to Read ID The full rectangle is visible with all corners intact, but OCR failed on specific fields, often due to glare, a worn laminate, or unusual fonts, not because part of the document is missing.
Text and photo look smeared or out of focus Document Blurry Corners and edges are all present, but the image lacks sharp focus, usually from camera shake or the lens focusing on the wrong distance. Cropping and blur are independent problems that can occur together or separately.

Let Us Help You Fix This

Try our free AI analysis first, or let our experts fix it for you.

Frequently Asked Questions

Why do all four corners of my ID need to be visible?

KYC systems use the corners to detect the document's full boundary, confirm its aspect ratio matches a known ID template, and check border security features like microprint and guilloche patterns that run close to the edge. Missing a corner makes it harder for the system to confirm the image shows a complete, unaltered document.

Does my phone's auto crop feature cause this?

Yes, this is one of the most common causes. Scanner-style camera apps try to detect the document's edges automatically and crop to that boundary, and if lighting, glare, or a cluttered background confuses the edge detection, it can crop slightly inside the real corners without you noticing until after you submit.

How much background should I include around my ID?

Aim for roughly 5 to 10mm of visible background on every side, enough that you can clearly see a gap between the document edge and the frame border. You do not need a huge margin, just enough that a slight auto crop or perspective correction still leaves all four corners intact.

Can I use a scanning app instead of the regular camera?

You can, but check the preview closely before submitting since scanning apps are exactly the type of tool most likely to auto crop your document. If you use one, disable auto crop if the app allows it, or manually widen the crop boundary before saving the final image.

Why does my document get rejected as cropped when I can see the whole thing on my screen?

What you see in a photo app's preview is sometimes the uncropped original, while a separate auto crop or compression step trims it before upload. The version your phone displays and the version that actually gets submitted are not always identical, which is why checking the uploaded file itself matters more than trusting the camera preview.

Other Common Errors