Automating Google Workspace (GWS) & more with Apps Script
Location and Vision Based Augmented Reality on Android
1. Location and vision
based Augmented
Reality on Android
Mostafa Akbari & Simon Heinen
1
09.04.2013
2. About us
• Working with AR since 2007
• First version of the DroidAR framework in 2009
• First presentation on the first droidcon
• Founded bitstars in 2012
• Startup out of the university RWTH Aachen
• Augmented reality and location based service infrastructure
• Framework for location based services and crowdsourcing
2
09.04.2013
3. Mobile Developer's Guide To
The Galaxy
• Implementing Augmented Reality
• p.171 AR Usage Scenarios in Mobile
• p.173 Tracking
• p.174 Augmented Reality SDKs
• p.177 AR Developing 101
3
09.04.2013
4. DroidAR v1
• Open Source since 2011
• github.com/bitstars/droidar
• Location based and marker based AR
• Project driven development
• 5000 developer worldwide
www.youtube.com/watch?v=brlNkKmw8IY 4
09.04.2013
10. More DroidAR v1 Examples
Example videos at 10
youtube.com/user/bitstars/videos
09.04.2013
11. DroidAR v2
JMonkey
Engine
Sensor and
Vision Based
Position
composition
system
11
12. jMonkey Engine
• Open source & cross platform
• ~358 years of effort (COCOMO model)
• First commit in June, 2003
• Test the application/game on the development machine using
mockups for Android specific components
• SDK based on NetBeans
• New BSD License
• Large community
12
09.04.2013
13. Developing your AR app
Create the game/app in jMonkey
Test it on the development machine
Create an Android app and add the
DroidAR framework
Add interfaces for Android specific
interaction
Use the DroidAR game loader in your 13
Android app
09.04.2013
17. SLAM
• SLAM = Simultaneous localization and mapping
• No dependencies to environment and no pre-knowledge
needed
• Very accurate positioning possible
• Creates 3d point cloud of environment
17
09.04.2013
19. Possible extensions
• Detecting surfaces
• Positioning objects
• Occlusion and shadow casting of virtual objects
• Barriers for physics simulations
• Object reconstruction
• 3D Object detection
• Matching the recorded features to a pre known 3d object feature
set
19
09.04.2013
20. Position composition
• SLAM
• works well indoor
• Other feature distances in outdoor environments
• Lighting conditions
• Absolute positioning components
• GPS
• WiFi
• Relative positioning components
• SLAM
• Step detection
• Composition system
• Weighting of positioning components 20
09.04.2013
21. Conclusion
High flexibility Easy to use
• Hybrid tracking: Automatic • Reusable components
composition system for positioning • Fast development process
and orientation • No technology lock-in
• Game engine approach • Combinable with our backend
• Targeting high scalability scenarios solutions for location based
applications
Scenarios Customers
• POI Browsers • Tourism
• Games • Education
• Navigation • Industry
• Multi-user/-player scenarios • Culture
• Instruction guides • Game developers
21