Tracking the users position and registering objects with high accuracy even on standalone devices is essential to get a convincing experience of AR applications. In this webinar, a state-of-the art review, major current challenges and an outlook to the future is given. The following experts will kick-off the discussion with stimulating talks:
Jérôme Royan, b-com, France
Tracking and registration for AR: How it works, what are the remaining challenges and new trends?
A good Augmented Reality system requires robust and accurate tracking and registration in all conditions in order to provide the user with a good quality of experience. The talk will cover the following points:
- A better understanding of what registration, tracking, mapping and relocalisation are, by comparing the functioning of AR systems with that of the human vision system.
- What are the limits of an AR system compared to the human visual system?
- How could AR systems soon surpass the capabilities of the human visual system thanks to machine learning and cloud computing?
Uli Bockholt, Visometry, Germany
Advanced Computer Vision for industrial Augmented Reality – Why understanding tracking deep tech is still so essential for enterprise use.
When it comes to industrial augmented reality, advanced computer vision technologies are key in order to enable a sophisticated AR at all. Tracking makes or breaks a use case -as such, understanding core tracking techniques, potential and limits, is key to AR’s successful adoption in enterprise use. Exemplary Application field is Augmented Reality Quality Control. Herey, the tracking has to fulfil the following requirements:
- Model-based Tracking Initialisation: The object captured with the camera has to be automatically aligned to the CAD model being its virtual counterpart.
- Moving Object Tracking: The AR-Inspection system has to support tracking of moving objects as otherwise with every transformation the system has to be reinitialised.
- Multi-Object-Tracking: AR-Measurement between two related components can only be realized if multiple objects captured with the camera can be tracked independently from each other.
- Multi-Camera-Tracking: Highest reliability for Computer Vision based Inspection can be reached by using synchronized camera arrays for tracking (e.g. installed on a production line).
In this talk, we introduce essential tracking techniques, we derivate tracking requirements for specific industrial AR-Use-Cases and we leverage Computer Vision beyond being the tracking foundation for simple graphical AR augmentations.
- Muriel Deschanel, Chairman of ETSI ISG ARF, Member of XR4ALL Consortium, bcom, France
- Oliver Schreer, Fraunhofer HHI and coordinator of XR4ALL, Germany
More info about the series of online events organized by XR4ALL, ETSI and The AREA can be found on the ETSI website.