Master and Bachelor Projects

Tabletops and Public Displays

CampusMap

CampusMap is a tabletop application to inform students about lectures, bus departures and the canteen menu. A custom-built kinect framework tracks multiple users around a tabletop to enable personal spaces.

Interactive Mirror

A graphical user interface for an interactive mirror. The mirror allows people to choose and wear virtual clothes and accessories in real time in front of a wall mounted multi-touch display.

Appstarter

The Appstarter is an interactive tabletop application written in LibAVG. The appstarter allows to start, manage and control applications that use LibAVG or the TUIO protocol.

SheeTra

SheeTra allows for tracking of physical sheets on multi-touch tables. We implemented two different trackers: A markerless tracking using the infrared camera and a markerbased tracker.

GeRSy

GeRSy is a gesture recognition system for multi-touch tables. For demonstration purpose a plugin for Google Sketchup allows navigation and manipulation of objects in 3D space.

Mobile Applications

Automatic Correction of Keystone Effects and Hand Jitter

Mobile projectors solve the limited display and interaction space on small mobile devices. This thesis combines a mobile projector with a Wii Remote to reduce keystone effects and hand jitter in the projected image.

LENS: Locate and Explore Nice Spots

LENS is a mobile recommender for Android to help photographers to find nice spots. A server application aggregates similar Flickr pictures into spots and compares them with the tags in the profile of the users.

DigitalJ

The project DigitalJ allows its users to connect and interact with an event. The application shows playlist and event information to the user and allows to provide feedback to the DJ, to vote and to wish songs.

Visualizations

Ray Tracing

In a rendering competition we implemented six ray tracing techniques to render a self-designed scene. All objects in this brewery scene were modeled using Blender and 3ds Max 2010.

Time-Series Descriptor

Time series provide massive data of multiple input sources. The implemented descriptor helps to reduce the amount of data, allows better comparison and visualization to the user.