# How to Scan an ID Card With a Phone: Both Sides, One Page

- Author: Schema Law (https://www.schemalaw.com/)
- Published: 2023-12-10T06:46:24Z
- Updated: 2026-10-11T19:22:20Z
- Category: Phone scans and photos
- Canonical: https://www.schemalaw.com/blog/how-to-scan-an-id-card-with-a-phone

> How to scan an ID card with a phone: capture front and back in one session, put both on a single page, read the small print, and store it safely.

To scan an ID card with a phone, capture the front and the back as two images in one scanning session, each filling the frame, on a dark matte surface with the flash off. Save them as one PDF. If the recipient wants both sides on a single sheet, place the two images on one page afterwards.

## Can a phone scan both sides of an ID card onto one page?

A phone can capture both sides of an ID card in one session, and the built-in scanners save them as two pages of one document. Putting both sides on a single sheet is a layout step you do afterwards, or an option to look for in the scanning app you use.

The documented behavior of the built-in tools is page by page. Apple's instructions for [scanning documents on an iPhone or iPad](https://support.apple.com/en-us/108963) end each capture with a choice: tap Done "or add additional scans to the document". Google's steps for [scanning documents with Google Drive](https://support.google.com/drive/answer/3145835?hl=en&co=GENIE.Platform%3DAndroid) on Android include "Scan another page: Tap Add" and let you save the result as .pdf or .jpg. Neither page describes a mode that arranges a front and a back on one sheet.

So you have three ways to reach "one page":

| Route | How | Good for |
| --- | --- | --- |
| Two-page PDF | Scan front, add the back, save | Your own file; conversion to text |
| One sheet, laid out by you | Insert both images into a blank word processor page, front over back, and save as PDF | A recipient who asks for a single page |
| An app's ID mode | Check whether your scanning app offers a two-sided card layout | Regular intake of many IDs |

Ask the recipient which they want before you start. A request for a "copy of ID, front and back" does not say which layout is expected. The two-page PDF keeps each side at its full captured size.

## Photographing the front and back of a license or passport page

Photograph each side separately, with the card or page filling the frame, and take the back even when it looks empty. Use a dark, non-reflective surface, light from the side and no flash. For a passport, capture the whole data page including the lines of characters along its lower edge.

1. **Set the surface.** A dark placemat, a book cover or a sheet of black paper. The card's pale edge then stands out and the scanning app can find it.
2. **Kill reflections.** Turn the flash off. Stand so no ceiling light sits directly over the card. The post on [how to photograph a laminated document](https://www.schemalaw.com/blog/how-to-photograph-a-laminated-document) covers glare from coatings in full.
3. **Get close.** Bring the phone down until the card nearly fills the screen with a thin border all round. If the picture goes soft, lift the phone slightly until it sharpens.
4. **Hold square.** The card's edges should run parallel to the edges of the screen.
5. **Capture the front.** Check that every corner is in.
6. **Turn the card the same way each time.** Flip it left to right, as you would turn a page, so the back is the right way up next to the front.
7. **Capture the back.** Barcodes, restrictions, endorsements and the cardholder's signature may sit there.
8. **Zoom in before you finish.** Read the smallest line on each side from the phone's screen.

For a passport, hold the book open with fingers at the outer edges only, well clear of the print. The booklet will want to close, so press the opposite page flat with your other hand outside the frame. If the facing page carries a signature or an observation, photograph the full spread as a third image.

A worked example. Priya Raman needs a copy of a new client's driver's license at intake. She lays it on the dark cover of her notebook beside a window, scans the front in the phone's document mode, adds the back as a second scan, and saves one PDF named with the matter number. Before handing the card back she zooms in on the license number and the expiry date in both images.

That PDF is a photograph of an ID gathered for an intake form. Read the license number off the image before you trust any text you export. Both sides, kept at full size, can stand as documentary evidence of the card that day. The matter name plus the date give authentication and the chain of custody a short trail.

## Why is the small print on an ID hard to capture?

The small print on an ID is hard to capture because the type is tiny, it sits on patterned and reflective security backgrounds, and the card is small enough that people photograph it from too far away. Each of the three takes detail away from the characters.

**Distance.** A card about 3.4 inches wide that fills a frame 4,032 pixels across is recorded at roughly 1,190 pixels per inch. The same card photographed from the height you would use for a letter-size page, so that it occupies a third of the frame, gets about 400. Hold the card next to a typed letter and compare the lettering. Where the card's type is the smaller of the two, the card needs the closer shot.

**Background.** Look closely at the card. Where text is printed over fine line patterns, a tinted portrait or a color gradient, a person still sees the text stand out. To software deciding which pixels are ink, the pattern is competition.

**Coatings and holograms.** A laminate reflects light as a mirror does, and holographic overlays flash bright at some angles and vanish at others. A reflection that falls across the date of birth erases it from the picture. Tilt the card a few degrees or move the light until the patch slides off the text, and take a second image if one angle cannot clear every field.

**Raised and engraved characters.** Some cards carry embossed or laser-engraved numbers with little color contrast. Light from a low angle at one side throws small shadows that make them readable.

Check your result against this short list before the card goes back to its owner:

- full name, every character
- document number
- date of birth, issue date, expiry date
- address, if printed
- anything on the back in small type

## What will OCR read from a photographed ID card?

OCR reads the printed characters it can see on a photographed ID: names, dates, numbers and labels. It returns them as text in reading order, not as labelled fields. Barcodes are not text, and the coded lines on a passport page come out as a string of capitals and symbols that needs checking character by character.

### Printed fields

On a sharp, close image the plain printed lines convert like any other print. Expect trouble where characters look alike. Document numbers mix letters and digits, so O and 0, I and 1, S and 5, B and 8 are the pairs to compare against the card. Field labels printed in very small type beside each value may be dropped or run into the value.

### Machine-readable lines and barcodes

Look at the data page of the passport in front of you. Along the lower edge there are lines of capital letters, digits and "<" symbols in an evenly spaced typeface, commonly called the machine-readable zone. General OCR treats them as text. Because the lines have no words in them, nothing in the context helps the software, and one wrong character changes a number. Read them against the image.

A barcode on the back of a card is a pattern, not lettering. Text recognition will not turn it into the data it encodes.

### What to do with the output

Schema Law reads uploaded pages into text and tables and exports them in any of its eight formats, on every plan. It does not pull out a "date of birth" field for you, so treat the export as a transcript to copy from, with the image open beside it. Recognition is strongest on clean print, and nothing in the result marks a doubtful character.

Before any identity document goes to an online tool, decide whether it needs to. A license carries several identifiers in one place. The post on [OCR for documents with Social Security numbers](https://www.schemalaw.com/blog/ocr-documents-with-social-security-numbers) covers the upload decision and the screening routine for identifying numbers. If all you need is six fields for an intake form, typing them by hand from the image leaves fewer copies.

## Where an ID scan should and should not be kept on a phone

An ID scan should stay on a phone only as long as it takes to move it to the firm's own storage. It should not sit in the camera roll, a personal cloud library, a shared album or a message thread. Scanning into a files location you control, then deleting, leaves the fewest copies.

Where the image lands depends on how you captured it.

| Capture method | Where the copy sits | Concern |
| --- | --- | --- |
| Camera app | Photo library | May sync to a personal cloud account and to other devices |
| Scan saved to a files folder | The folder you chose | Easier to find and remove |
| Scan inside a notes app | The note | Easy to forget; syncs with the notes account |
| Sent by text or chat | The thread, on both phones | Stays until each person deletes it |

Deleting from the photo library is not instant removal. Apple's page on [deleting photos on your iPhone or iPad](https://support.apple.com/en-us/104967) says deleted items "go to your Recently Deleted album for 30 days", and that with iCloud Photos a deletion on one device applies to your other devices. Google's page on [deleting photos and videos in Google Photos](https://support.google.com/photos/answer/6128858?hl=en&co=GENIE.Platform%3DAndroid) says deleted items "stay in your trash for 30 days before they are permanently deleted". The post on [deleting scanned documents from a phone](https://www.schemalaw.com/blog/delete-scanned-documents-from-phone) walks through every place a copy can remain.

A short routine for staff who capture IDs on their own phones:

1. Scan to a files location, not the camera roll, where the phone allows it.
2. Move the PDF to the matter folder the same day.
3. Open the copy in the matter folder and confirm both sides are there.
4. Delete the phone copy, then clear it from the deleted-items area.
5. Delete any message or email draft that carried it.

When a client sends their own ID by text, the copy on the client's phone is theirs to manage, but the one in your thread is yours. Save it to the matter and remove it from the conversation. If you do upload an identity document for conversion, read what the service commits to first. Schema Law lists its commitments, including "Delete any document whenever you choose", in the [security section of the homepage](https://www.schemalaw.com/#security).

## Try it on a real document

Upload a scanned PDF and export it as Word, a searchable PDF or plain text on the free plan. Your documents are never used to train an AI model. [Start free](https://www.schemalaw.com/api/auth/google)