How to Prepare Images AR Recognizes Well: A Pre-Print Checklist

In augmented reality, the phone camera recognizes a printed image and plays the video linked to it. The faster and more reliable the recognition, the more magical the experience feels. The good news: recognition depends mostly on the image itself, and a few simple rules let you get it right before you print.

How does the camera recognize an image?

The recognition system looks for distinctive points in the image, such as corners, edges and texture, and remembers how they are arranged. When the camera looks at a page, it tries to find the same arrangement. That's why detailed, textured, high-contrast images are easy to recognize, while flat-colored, empty or repetitive images are hard.

Pre-print checklist

1. Is there enough detail?

Photos, illustrations and pages that mix text and graphics are usually recognized well. Large flat color areas, sky or mostly white backgrounds and very minimal images are hard to recognize. If much of your image is empty, consider using a more detailed part of the page as the trigger.

2. Is the contrast good enough?

In very dark, very light or washed-out images the distinctive points get lost. Mid-tone images with clear light-dark transitions work best.

3. Are there repeating patterns?

Repeating patterns like tiles, fabric prints or brick walls can confuse the system, because the same points appear everywhere. If you're showing a patterned product, including some of its surroundings in the photo helps.

4. Do the pages look too similar?

Pages built on the same template where only a small product photo changes can get mixed up. Give every trigger image something unique: a different photo, color or composition.

5. Is the uploaded image the same as the print?

The image you upload should be the printed page itself: the image from the design file or a flat, sharp photo of the page. If you cropped or changed colors after printing, update the uploaded image too.

6. Is the resolution high enough?

The uploaded image should be at least a few hundred pixels on its short side; under 480 pixels is weak. The system checks each image as you upload it and warns you if it looks small, dark, too bright or low in detail.

7. Is the surface glossy?

Glossy lamination and varnish can reflect light and “erase” part of the image. Choose a matte finish where you can; the difference is especially noticeable on menus and posters.

8. Do print size and scanning distance match?

The camera recognizes an image when it fills most of the frame. Large posters are scanned from a few steps back, small catalog images up close. A very small printed image (stamp-sized, for example) is hard to recognize.

9. Is the ambient light good enough?

Cameras need light. Dim restaurants, evening exhibitions or backlit posters can make recognition harder. Test the image in the environment where it will be used.

10. Did you invite the reader?

Even the best-recognized image is useless if nobody scans it. Put a short invitation and a QR code on the cover, the table or a corner of the poster: “Scan with your phone and watch the video.”

Final step: test

Before sending to print, take a few minutes to test:

1. Upload the images and videos.
2. Scan them on screen or on a test print with different phones.
3. Try again under lighting similar to where the image will be used.
4. Replace any image that is hard to recognize with a more detailed photo.

These rules are the same for catalogs, exhibitions, menus and posters. For examples, see the product catalogs, exhibitions, menus and posters pages.