Adaptive Content Harvesting, Creation and Delivery

This workpackage deals with the ‘user-oriented data’ used to power the VENTURI platform covering issues such as ‘data flow’ (user needs driving the sourcing, harvesting, searching and processing), and the research and development concerning adaptive content creation and delivery methods for virtual objects to be placed into the real world,  producing methods for 3D visual reconstruction/insertion and visual content fusion.
Outcomes are therefore:

  • Off-line 3D reconstruction methods from multiple views captured in a lab environment using a calibrated multi-view setup, and also from other image sources (historical photographs or paintings). The outcome of this task will be a sophisticated framework for 3D reconstruction from different visual sources
  • Fusion of the 3D geometry of the virtual objects with other data, e.g. images or video data. Here, novel methods for multi-modal registration (e.g. 3D shape and images) will be developed
  • Large scale Web-of-Data social content mining. This refers to the process by which location and sensor information are used to trigger searches for relevant pieces of information on the Web-of-Data, thus the system can be ready for certain events and objects that are likely to be present in the surroundings.

Finally, the virtual objects have to be transmitted and delivered to the user through the mobile portal. An API to handle the delivery will be developed.

Comments are closed.