Augmented Reality can be provided through a smartphone camera or
an AR headset.
AR requires tracking of real world targets. These targets can be either the world (i.e.
the camera’s orientation and location in the world), an object, an image, marker, face
or entire body. Here’s an outline of the types of tracking you may encounter:
SLAM: tracks the camera’s orientation and location in the world.
Plane tracking/detection: is able to find planar surfaces in the world for positioning
Marker tracking: finds black and white marker images to attach virtual objects to or
track the camera’s location and orientation in the world
Image tracking: finds images to attach virtual objects to these images, e.g. through a
process called “natural feature tracking”.
Face tracking: is able to track features in a human’s face to attach virtual objects to.
Full-Body tracking: is able to track entire human poses, e.g. for virtual clothing.
Some other terms you might often hear when talking about AR:
Light estimation: the process of detecting light sources in the real world, for replication
on virtual objects, which can greatly increase realism and the perception of the object being
integrated in the real world.
Depth occlusion: when the camera is able to detect how far objects in the real world are away
from the camera, this information can be used to hide virtual objects such that they appear to be
behind real world objects.
Anchor: a “pin” in the real world that can be used to attach virtual objects that then appear
to stick in the real world.
Hit test: process of “shooting” a virtual ray from the camera into the real world to check the
distance of the real world to the camera. Often used to place virtual objects.
Smartphone AR can be achieved in two ways:
the WebXR API is not available.
WebXR Device API - AR Module: is natively implemented in the browser, therefore
light weight and performant. SLAM is supported,
Wonderland Engine comes with an integration with the 8thwall AR tracking
framework, which provides SLAM tracking, but also other AR tracking methods like image- or face tracking.