Home > Teaching > Theses > Context-aware Photo Sharing

Context-aware Photo Sharing

Type: Bachelor thesis
Status: Finished
Begin: From March 2014
Student: Réne Jahn
Tutor:

Research Area

Today’s powerful mobile devices and software platforms provide exceptional foundations for wide-area context-aware computing; however, existing context-frameworks are often difficult to use, lack appropriate security features, support few context types, and are unable to integrate new capabilities at runtime. To address these challenges, we’re developing Ambient Dynamix, an open plug-and-play context framework for Android. Dynamix runs as an OSGi-based background service on Android-based devices, modeling context information from the environment using the device itself as a sensing, processing and communications platform. Context modeling is performed by a tailored set of plug-ins, which are dynamically provisioned to the device during runtime. Plug-ins relieve application developers from the complexity of low-level context modeling issues, such as sensor management, noise reduction, time-of-flight calculations, feature extraction, machine learning, web-services interaction, etc. Dynamix comes with a large collection of ready-made plug-ins and provides an Open Plug-in SDK and extensible repository architecture, which enables 3rd party developers to quickly create and share new plug-ins types with the community.

The Thesis Topic

The ubiquitous availability of digital photo and video cameras has created an enormous increase in daily life photo shooting session. Especially with the almost zero costs involved, people take plenty of photos during their daily activities. Still, the majority of pictures are captured at touristic scenes. The main problem of mobile photo shooting is nevertheless still the fact, that you can't be photographer and photographed subject at the same time. Therefore, many pictures are taken of touristic places with many unknown persons in the foreground or background of the scene, except yourself.  

In this thesis, a novel photo sharing application should be developed, that combines the domains of social collaboration and the dynamic context acquisition features of Dynamix. Main idea is to allow people to share their pictures to nearby people in crowded areas based on proximity parameters, like similar location, time, devices in range, etc. This would allow to include the relevant pictures from the many taken at the same event.

This thesis should:

  • Investigate appropriate proximity parameters and design a "nearby"-Function.
  • Develop necessary Dynamix-PlugIns if needed.
  • Investigate the use of face recognition tools to automatically (pre-)select relevant pictures.
  • Design and implement a prototypical Dynamix App in Android.
  • Investigate the integration of state-of-the-art cameras into the workflow (e.g. WLAN upload to smart phone, GPS meta data, etc.)
  • Perform an evaluation of system performance.
© 2016 University of Lübeck · Institute of Telematics.    All rights reserved.