Special considerations for Standards in Mobile AR. International Workshop on AR Standards – 11. + 12. October 2010, Seoul [email protected] - Institute for Graphics Research IGD. Fraunhofer IGD Institute for Computer Graphics Research. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Special considerations for Standards in Mobile ARInternational Workshop on AR Standards – 11. + 12. October 2010, [email protected] - Institute for Graphics Research IGD
The aim of the presentation is triggering consciousness about the idea of AR and presenting an open approach practically ..
Where are the problems in standardization of AR? What are the common structures and where are variations? What is the role of those structures in the context of AR? Technological hardware issues Why and what kind of standards are there and do we need? Our approach to overcome the mentioned issues and help
developing standards
Cannot cover the whole horizon of problems .. But can try …
Retrieval of the camera Pose (6Dof) Find the absolute coordinates of user and objects
GPS, Compass, Acceleration Data Use prototype model for registration (i.e. Poster, Marker, Lines…)
Representation Absolute in latitude, longitude, altitude, roll, pitch, yaw Relative in Cartesian coordinates and angles (or axis angle, quaternion..) In relation to some object or object group – real or virtual?
Interaction Find differences of formulated and real world (POI,AOI,VOI)
This creates potential content for interaction Track movable and deformable objects Detect user interaction/behavior using sensors (Hands, gestures etc..)
Customizing the Annotation Model Manipulation Scene Graph Interactive Content
Customizing the Reference Model Dynamically change tracking behavior and point of focus (POI) Allow interchange in between Annotation and Registration Model Use of declarative formulation as far as possible. Allow scripting for specialization Tracking and Events: 2D-Projected Positions with distance hint Detection Events (i,e. Object enters/leaves display) Interaction of user/objects with the scene
Content must be delivered free or encrypted.. Who will do this?
Low computational resources in relation to modern PCs Limited memory OpenGL ES Different camera modules (Intrinsics) Sensor support and calibration Partly supported hardware acceleration Software Interfaces and Languages Symbian (C++, Microkernel, Qt) Android (Java, C++ through NDK) iPhone (Obj-C, C++ support)
Algorithms need to be optimized for HW Costly and hard to maintain large set of handsets in one common
Open StandardsSustainability for AR-ContentApplications and generation tool chains should be based on.. X3D (ISO/IEC 19775:2004) HTML5 ECMAScript (ISO/IEC 16262:2002) JavaScript CSS STEP-Format KMLBenefits Sustainability Interoperability Porting Rapid development
X3DOM (x3dom.org)HTML5/X3D Integration ModelX3DOM: experimental open-source framework and runtime One way for integration of HTML5 and declarative 3D Ongoing discussion in the Web3D and W3C communities
Declarative content design Allows embedding 3D data directly into the HTML-DOM First approaches of CSS integration Application concepts map well to generic scene-graph
Runtime independent (supports native implementation, plugin, WebGL)Suitable for desktop and mobile systems alike
mobileAR – AR – Browser for mobile devicesAdditionally to Web-Browser a renderer and vision engine is integrated. Extends the internal Web-Browser with a simple but powerful API Interaction and visualization can be done declarative in HTML and/or
via JS Can render X3D via native renderer or even X3Dom Generic Vision Interface Integrates Vision Processor (VisionLib) for
Fiducials (Markers, Image-Markers) Poster-Tracking Object recognition Panorama and 3D-Model Based scene tracking Server based Tracking (bidirectional streaming)
Additional Features (about to come…) Remote Video & Audio Support Scenarios Encryption and packaging
Next steps…Identification and unification of common technology data Annotation Data Reference Data
Create open framework for declarative approaches Open Annotation Engine Open Context Engine Unify Vision and Sensor based approaches KML in the context of industrial maintenance applications Evaluate using show-cases