Augmented Reality is a delightful technology, but even the best intentions can go wrong if you don’t use the right images for your experiences. This is why CraftAR takes the guesswork out of AR. We provide an extended Image Quality rating for your items, not just in the cloud content management system, but directly on the user’s device, too.
This is very useful in case you want to 
let your users create their own AR experiences, for example, on top of their photos or print images. Now, you will be able to ensure that they will work properly.
Intro: creating Augmented Reality experiences with CraftAR
Our CraftAR toolbox and AR SDKs allow you to integrate Augmented Reality experiences into your apps. Whenever creating an experience, you first need to define a reference image.
Such reference images represent real-life objects, such as a print photo, an ad in a magazine, a poster, etc.
You will be able to ‘overlay’ digital content such as 3D models or a video on your reference images. You may develop a way to do it inside your own systems, but you can also use our drag-and-drop AR content creator tool in CraftAR. The content will appear on top of the object inside the camera view whenever the user scans the object with the app in real life.
You can add, access or remove reference images for AR purposes in different ways. For example, via our Management API, dynamically created inside your app – see user-generated content – or uploading them to the CraftAR web panel.
But how do we know whether our reference images are good enough for a stable Augmented Reality tracking?
How CraftAR’s Image Quality Rating for AR can guide you
When viewing an item in your collections in the CraftAR platform, you will see an Image Quality field with an orientative 5-star rating. This indicates how well that specific image is supposed to work in normal conditions when used as a reference image for AR, assuming that the object is flat, see a poster, banner, print photo, magazine etc. As a rule of thumb, above 3 stars, it will probably work well.
Our Augmented Reality solutions rely on the texture in the image – the visual elements – for tracking. This is why the best quality scores are for those items that have a detail-rich, non-repetitive visual pattern all over the surface of the image. This ensures that the AR tracking can be precise and stable.
Let’s see some examples for high and low Image Quality scores:
The beach photo reached a relatively low Image Quality score for AR purposes.
The main reason is that the pattern is not very distinctive and it features large, almost textureless areas.
On the other hand, this photo, depicting a mural, has a unique, detail-rich pattern,
so it will work much better for AR tracking.
 |
 |
The interior design photo on the right would serve quite well as an ad in a magazine
that users can scan for an AR experience.
The photo of the plain white chair in front of a white background has much less distinctive features across the image,
so it reached a lower score.
Check out some further tips regarding how to create your collection of reference images for AR experiences.
User-generated content for Augmented Reality experiences
Gone are the days when mobile users were only able to consume Augmented Reality experiences. Nowadays, more and more apps emerge that allow users to create and share their own AR content.
Prynt used our CraftAR SDKs to let people turn their print photos into Harry Potter-style moving images, recording a short video when taking the photo and unlocking it as AR content with the app.
Or, see the Tagxy app that offers a ‘social treasure hunt’ with user-generated content and AR experiences we create for posters, packs, places, magazines around us.
It’s plain to see why “Generation Me” enjoys participating in the creation instead of settling for the role of a spectator.
But how can app developers ensure that such user-generated content will work for AR tracking?
New: being able to evaluate Image Quality on the user’s device
With the latest version of our CraftAR Augmented Reality SDKs, you can now check the image quality for any image without the need to connect to our service. The computation happens on the device and returns the quality value.
As we’ve mentioned, this is extremely useful in cases where the final users generate the reference images and the content. With this, they can be guided on whether an image is good or bad for AR before creating an Augmented Reality experience, directly from the app. (Check out the documentations for iOS and Android.)
 |
 |
Two mobile photos taken on the same ski holiday, but one of them would work better for their AR experience.
Now you can help your users use the right images.
If you want to allow your audience to generate content, while maintaining a flawless user experience, then this is for you.
See for yourself. Download our free Augmented Reality SDKs with the advanced Image Quality rating options and try our CraftAR service.
—
Photo credits: Mashable / Prynt, The Odd Chair Company, Tiny Ski Lodge , West Elm