Augmented Reality (AR) is studied in this thesis from a desire
to deploy AR products in large scale within a few years.
This assumes a plain software solution, using Commercial
Off The Shelf (COTS) hardware as existing laptops, tablets
and phones. The thesis scope is a mirror metaphor, the device
mimics a mirror and AR is presented in the mirror.
The mirror is from looping video from the camera facing
the user to the screen and real-time AR graphics is added
to the video stream. A video mirror is slightly different
from a physical mirror and the consequences in this context
The more obvious video mirror from operating a device
in selfie-mode – recording a video of oneself – is recognized
to be used for AR features in large scale without being
marketed as an AR product. This is seen as sign of AR
maturity in the discussion.
A less recognized video mirror, by looping video from a
work area on a physical desk and apply AR on objects in
the work area is studied. The focus is the potential of using
a device’s face oriented camera to create an AR mirror
from unmodified COTS hardware. Even if AR is a well researched
area, the mirror metaphor by video loop has been
almost neglected. This can be explained by a) physical mirrors
are generally better for AR, especially in lab situations
or unique equipment, b) the AR research field matured long
before common COTS screen devices had cameras and c)
AR in mirrors has less immersion than AR in a direct view.
The basic properties of an AR implementation in a laptop
and a tablet are studied, obstacles are found and possible
ways to handle them are presented. The general recommendation
though is to add a USB camera to the laptop
or a tiny mirror to the tablet. Literature research indicates
that AR using the mirror metaphor offers a noticeable improvement
compared to screen or paper based presentations
when used in relevant situations.
It is concluded that AR is maturing. There will not be
an AR revolution but we are in the beginning of an evolution
where AR is included for its ability to improve user
interaction rather than for being impressing technology.
2016. , 48 p.