How to extract text from an image or photo

By Nasir Uddin Shamim · Founder, PrintxPDF

Updated 2026-07-31 · 6 min read

This used to require software. Now every major operating system reads text out of images for free, usually without you installing anything, and the results on a decent photo are close to perfect. The trick is knowing where each platform hid it.

The fastest method on each platform

Built-in text recognition, as of 2026.

Photographing a page so it actually reads

Most failures are capture failures, not recognition failures. Two minutes of care at the photo stage saves ten minutes of correcting text afterwards.

1. Get even light Near a window, no flash, no direct sun. Uneven light produces a bright half and a dark half, and the dark half thresholds to solid black.
2. Keep your shadow off the page Stand so your body and the phone are not between the light and the paper. This is the single most common cause of a failed capture.
3. Hold the phone parallel to the page Shoot straight down. A tilted photo gives you keystoned text where lines converge, and recognisers handle that badly.
4. Fill the frame Get close enough that the text fills most of the image. A capital letter should be at least 25 to 30 pixels tall in the final photo.
5. Tap to focus, then check Focus on the text, not the page edge. Zoom into the photo afterwards and confirm the letters are crisp before you walk away.
6. Use a document scanner mode if you have one The Notes app on iOS and Google Drive on Android both have a scan mode that detects the page edges, corrects perspective and flattens the lighting automatically.

Tip: If you are capturing several pages, shoot them all first and process them together. It is far faster, and consistent lighting across the set means one adjustment fixes every page.

Screenshots

Screenshots are the easiest case by a wide margin: the text is perfectly sharp, evenly lit and rendered at whole-pixel positions. Accuracy is normally at or near 100 percent, and even small interface labels come through.

The exceptions are text over busy photographic backgrounds, very low-contrast grey-on-grey interfaces, and text at 200 percent zoom that has been resampled by a screen recording. For a screenshot on a high-resolution display, capture at the native resolution rather than a downscaled copy.

When you have a folder full of images

Point-and-click methods stop making sense at about ten files. At that point, combine the images into one PDF and run OCR over the whole thing in a single pass.

1. Sort the files so the names put them in reading order, most tools use filename order, not the order you selected them.
2. Combine them with the images to PDF tool, which puts one image on each page.
3. Run the result through OCR once, choosing the correct language.
4. Export plain text with PDF to text if you want the words rather than the document.

Got more than a handful of images? Combine them into one PDF and OCR the lot in a single pass.

Privacy: on-device or in the cloud?

This distinction is worth knowing before you photograph a payslip. Live Text on Apple devices and PowerToys Text Extractor on Windows run entirely on the device. Google Lens and Google Drive OCR send the image to a server. Browser-based OCR tools that use WebAssembly also run locally, with only the language model downloaded.

Watch out: Photos carry EXIF metadata, which routinely includes the exact GPS coordinates where the picture was taken and the device that took it. If you are sharing a photographed document rather than just reading it, strip that first, the same reasoning as removing metadata from a PDF.

What comes out, and what does not

Recognition returns characters, not layout. Columns are usually merged into a single stream, tables lose their cell structure and arrive as loose runs of numbers, and hyphenated line breaks stay hyphenated. Expect to clean up. If the layout matters more than the words, keep the image and add a searchable text layer instead of extracting plain text, see how to OCR a scanned PDF.

Frequently asked questions

How do I copy text from a photo on an iPhone?

Open the photo, press and hold on the text, and drag the selection handles as you would in any document. This is Live Text, built into iOS since version 15, and it runs entirely on the device.

How do I extract text from a screenshot on Windows?

Open the screenshot in Snipping Tool and choose Text actions, then Copy all text. Alternatively install Microsoft PowerToys and press Win+Shift+T to drag a box around any text on screen.

Is there a free image-to-text converter?

Yes, and you probably already have one. Live Text on macOS and iOS, Text actions in the Windows 11 Snipping Tool, and Google Lens in Chrome are all free. Browser OCR tools handle batches without uploading anything.

Why is the extracted text full of mistakes?

Almost always a capture problem: uneven lighting, a tilted phone, or text too small in the frame. Reshoot with the phone parallel to the page in even daylight, filling the frame with the text.

Can I extract text from many images at once?

Combine them into a single PDF first, one image per page, then run OCR over the whole document in one pass. That is much faster than opening ten images individually and keeps everything in order.

Tools

• OCR PDF
• JPG to PDF

Keep reading

• How to OCR a scanned PDF so you can search it
• How to get better OCR results from bad scans
• How to combine photos and scans into one PDF