What Image Resolution Do You Need for Good OCR?
People assume that bigger images mean better OCR. That is only half true. What an OCR engine actually needs is enough clean pixels per character, with good contrast and sharp edges. Resolution matters, but it works alongside lighting and focus.
The rule of thumb: aim for 300 DPI
For scanned documents, 300 dots per inch (DPI) is the long-standing sweet spot. At 300 DPI, normal body text has enough detail for the engine to tell an "e" from a "c" and an "8" from a "B". Going much higher rarely improves accuracy and just makes files larger and slower to process. Going much lower, say 150 DPI, starts to blur fine strokes and accuracy drops.
When you are photographing rather than scanning, DPI is less meaningful. What counts is how many pixels cover each letter. A good target is roughly 25 to 30 pixels of height per character. If text in your photo looks crisp when you zoom in, you are usually fine.
Why a sharp small image beats a huge blurry one
A 12-megapixel photo sounds like plenty of resolution, but if it is out of focus or motion-blurred, those extra pixels are just blurry pixels. The OCR engine cannot recover detail that was never captured cleanly. This is why focus and lighting often matter more than raw size.
In practice, three things drive readability:
- Focus. Tap to focus on the text before you shoot.
- Lighting. Even, diffuse light with no harsh shadows.
- Framing. Fill the frame with the text so each character gets more pixels.
If you want the full checklist, our guide on improving OCR accuracy covers the capture habits that make the biggest difference.
When more resolution does help
Higher resolution genuinely helps with small print: footnotes, fine print on contracts, ingredient labels, or dense tables. If the original text is tiny, give the engine more pixels to work with by moving closer or scanning at a higher DPI. The same goes for reading blurry or low-quality images, where the limiting factor is usually clarity, not size.
Practical settings to use
- Scanning a document: 300 DPI, grayscale or black-and-white.
- Phone photo of a page: fill the frame, hold steady, good light.
- Tiny or dense text: move closer or scan at 400 DPI.
Once you have a clean image, just upload it to the Image to Text tool and the engine handles the rest. Files are auto-deleted after processing and there is no sign-up.
Common questions
Is higher resolution always better?
No. Past about 300 to 400 DPI you get diminishing returns: bigger files, slower processing, and little to no accuracy gain. Clarity and contrast matter more than sheer pixel count.
What is the minimum resolution that works?
There is no hard floor, but text smaller than roughly 20 pixels tall per character starts to fail. If you can barely read it on your own screen at 100 percent zoom, the engine will struggle too.
Should I upscale a small image before uploading?
Upscaling does not add real detail, so it rarely helps and can introduce artifacts. It is better to re-capture the source at a higher resolution if you can.
Get a clean read every time
You do not need a fancy scanner, just a sharp, well-lit image with enough pixels per character. Hit the 300 DPI guideline for scans, focus carefully for photos, and drop the file into the Image to Text converter. It is free and your upload never sticks around.