System for determining maskable and unmaskable background regions in an AR system (20-Jan-2010)
A concern with eyewear based augmented reality systems is that alerts and information may cover up important objects In Real Life (IRL). Examples of important objects may include emergency vehicles, trip hazards or open manholes. A solution is presented which describes a system whereby IRL objects and view regions are recognised and dynamically assigned levels of importance. Objects with higher levels of importance will not be covered by alerts and information displays for objects with a lower importance. Correspondingly, lower level objects and regions are coverable by higher level alerts and information displays.
System for determining maskable and unmaskable background regions in an AR system
A concern with eyewear based augmented reality systems is that alerts and information may cover up important objects In Real Life (IRL). Examples of important objects may include emergency vehicles, trip hazards or open manholes.
Simple solutions to this problem could involve the definition of a "centre" view region in which no information can be displayed. This does not allow for the situation where the user is focusing on an unimportant region, and a hazard appears from peripheral vision, yet is covered by information. This can also compromise one of the fundamental contextual benefits of AR where metadata is displayed close to the IRL object.
The solution presented herein describes a system whereby IRL objects and view regions are recognised and dynamically assigned levels of importance. Objects with higher levels of importance will not be covered by alerts and information displays for objects with a lower importance. Correspondingly, lower level objects and regions are coverable by higher level alerts and information displays.
Basic order of events:
1) The existing computer vision system uses existing visual recognition techniques to recognise and classify contiguous IRL regions (such as the sky) and objects (such a tree, person or car).
2) Our solution assigns each object/region a level of importance according to its class, and state (for example a moving ambulance with sirens on would have a higher importance than one that was stationary, or...
Find More Documents Like This
Download this Disclosure
- a representative PDF of the primary disclosure file (contains all the relevant information for most users)
- the full disclosure ZIP file containing the primary disclosure, packaged metadata, and attachments (as appropriate)
