When working with any new or existing technology – whether this would be HTML5, augmented reality, the Xbox360 Kinect, stereoscopic 3D or others – it’s important to know the technical limitations in order to assess the possibilities.
We’re happy to share some of our findings from our internal testing with augmented reality (AR) technology – the Qualcomm AR SDK for smartphones in particular.
A video of this technology in action can be seen in a previous blog post.
These are our findings:
- If you want your virtual objects to be displayed naturally in the real world, you’d have to set the light color and direction match these in the real world. There doesn’t seem to be an easy solution for this, but could be done by analyzing the images captured by the camera. When using marker images, a possible solution to get the appropriate light color may be to take the perceived color from white areas in the marker image and use this as the light color.
- When using marker images, you have to know that players can print these images at any scale they want – affecting the size of the virtual objects. This can be a problem for physics based games, as objects with different sizes behave differently.
- When a marker becomes visible, it takes some time until it’s recognized. This results in the 3D object popping up suddenly. One way this may be resolved or improved is by cropping the image from the camera when displaying it in the game, so the borders aren’t visible and the marker can be detected before it would be visible to the player.
- The detection of the markers can be slow, and virtual objects don’t properly follow when you swiftly move around your mobile device. This makes this implementation of marker-based augmented reality unsuitable for games where this kind of movements are required.
AR technology may trigger your imagination and you may come up with the wildest game concepts, but in the end it’s crucial to be aware of the technical limitations – otherwise it becomes very difficult to assess the feasibility of your ideas and to come up with solutions needed to realize these ideas.
3 thoughts on “Augmented Reality: Our Findings”
have you also tried String?
We haven’t tried String yet, but it’s likely that we’ll try it out in the near future.
Looking at their website, I see they have a solution to create more natural lighting (Live Color Sampling, as they call it), which could be very interesting.
Meanwhile, Qualcomm AR has been renamed to Vuforia, and has been acquired.